_ r , m. ~- $13..“ . an ’ \ v» " ' 1, ' :92? ant" "'75 m M" .1 M . “3-551?!“- (5651"; .1 All 4, ':L:"‘ :- iv v . . .71." r m»: 9 . '4 N'rf'l- 953;: .3 ‘ fi 35;}: .. ' ‘ «:7. AA“ a.“ mum '"g 24.." . :jvngurg . .- :N: aw?“ "' ' a ’fi ”“943? J W! I. ) "yr": «aw ' ' h 2’ 1 ‘ _ l f 1 v1.7 rw . ‘r- '1': r a 5:; £511, a“; I - J . w ‘ ',f... N .'.’ :j’n: yv‘ '- {10" 1 ~ ,...r.5 - 1...” cu 3"“- . ' J 11‘1”” "I, .IICIf emf- urn." " PM)” r. n." . .7 \ .. ,, WM p-v-L'u“ v-Iw win. ‘ ‘, r. .0 r‘ hr: 7:, m3.- 1’l ; ,- ft ".3,“ .v Ir _, v , -. ' ' :C':flu-’" v- n“ 5" ’ ‘ 1: u‘ ‘ . [l:“! 1' v nw...¢.,.; v r- 4 .- {p . . .‘,',. 53!...” a H. Y rm #:- - I «.1 or’ .' , t WAN.“ -‘ N»: .u-.. 7 .,.,.(/5 5:? gawk: 1‘. - #151,! ‘ 1;..1.“ ‘ 1 hi nLAH. .g. «L #4343 & JIllllllillllllllTlllllllllllll lll L 3 1293 00824 2343 This is to certify that the dissertation entitled IDENTIFICATION AND CHARACTERIZATION OF MAREK’S DISEASE VIRUS GENES PUTATIVELY RESPONSIBLE FOR IMMUNOPROTECTION AND ONCOGENICI’I'Y presented by Xinbin Chen has been accepted towards fulfillment of the requirements for DOCTOR OF PHILOSOPHY degree in Dept. of Microbiology and Public Health ajor professor Date /€/2;/'77 MS U is an Affirmative Action/El] ual Opportunity Institution 0- 12771 PLACE N RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or More die due. DATE DUE DATE DUE DATE DUE l I MSU II An Mflflndlve ACUONEQUII Opponmity Indittllon IDENTIFICATION AND CHARACTERIZATION OF MAREK’S DISEASE VIRUS GENES PUTATIVELY RESPONSIBLE FOR IMNIUNOPROTECTION AND ONCOGENICITY BY Xinbin Chen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1991 623‘55 3X ABSTRACT IDENTIFICATION AND CHARACTERIZATION OF MAREK’S DISEASE VIRUS GENES PUTATNELY RESPONSIBLE FOR IMMUNOPROTECTION AND ON COCENICITY By Xinbin Chen Identification of Marek’s disease (MD) virus (MDV) genes responsible for oncogenicity and immunoprotection is essential to understand the mechanisms involved. Through DNA sequence and cDNA analyses, an open reading frame (pp38-OR?) encoding 290 amino acids was identified in BamHI-H. Antisera against trpE-pp38 fusion proteins irnmunoprecipitated 38- and 24~1>>._.>> 932 pm: >>._.>>> tear: _t AM _ v .v 6 .v _ q . noose . TIIL A II n oxemsooo 59o: .v ooz>ei .v H 95 ._ mute amuse” 1v genes. 11: .1... 5593262. 8203.: 32>“: 00256 #A and #m ..v ..v 1v .lv H 2323: o. no? .2 003335 accesses a. A. ooz> as Alllllltl llllll A ooz> 3m 7“ Test... F 44 Slmlcleasemapplngoftanstz'lptsfi'ompartlywithinormarthe expanded regions of BamHI-D and 41. Since the expanded regions of BamHI-D and -H are the same (sequence alignment between the expanded regions of BamHI-D and -H shows 98% identity; data not shown), only probes from the EcoRI-a subfragment of BamHI-H were used. All the probes used have five copies of the l32-bp direct repeat. Figure 4A shows the probes used as well as the sites in the probes where S1 nuclease out. To map the 5’ ends of tanscript which initiate rightwardly within the expanded region of BamHI-H, probe 1, a 760-bp DraI-Banl fragment 5' end labeled at it BanI site by [7-32P]ATP, was used. Seven major RNA-DNA hybrids (from 100 bases to 600 bases) were protected from $1 nuclease digestion, indicating that the respective tanscript are heterogeneous and contain between one and five copies of the l32-bp direct repeat at their 5’ ends (Fig. 4B). Since there are five copies of the 132-bp direct repeat in the probe used, one can not determine whether some of the tanscript might contain more than five copies of the 132 base direct repeat at their 5' ends. To map 5’ ends of tanscript which initiate leftwardly within the expanded region of BamHI-H, probe 2, a 760-bp DraI-BanI tagment 5’ end labeled at it DraI site by [7-32PJATP, was used. Two RNA-DNA hybrids (150 bases and 280 bases) were protected, indicating that the respective tanscript contained one or two copies of the l32-bp direct repeat at their 5’ ends (Fig. 4C). To map the 3’ ends of tanscript which terminate leftwardly within the expanded region of BamI-lI-H, probe 3, a 1.0-kbp SmaI-Banl fragment 3’ end labeled at it BanI site by [a-32PldCTP, was used. Ten RNA-DNA hybrids (from 170 bases to 750 bases) were protected, indicating that the respective tanscript contained between one and five copies of the 132-bp direct repeat at 45 FIG. 3. Northern blot hybridization to detect tanscript from the expanded regions. poly(A)+ RNA (0.5 pg) was used per lane. (A) Probe: 1.6- kbp SmaI-PstI subfragment which spans the expanded region of BamHI-H. (B) Probe: cDNA 3. (C) Probe: cDNA 1. (D) Probe: cDNA 4. (E) Probe: cDNA 11. (F) Probe: cDNA 15. INF, RNA from DEF cells infected with MDV. CON, RNA from uninfected DEF cells. The precise location of the probes is shown in Fig. 2. — 73°C 7 a: '54.). Li) LQ'O '- 1'8- 112- zoo 1'12- 200 112- 200 112- 200 112- 200 45 FIG. 3. Northern blot hybridization to detect tanscript from the expanded regions. poly(A)+ RNA (0.5 pg) was used per lane. (A) Probe: 1.6- kbp SmaI-Pstl subfragment which spans the expanded region of BamHI-H. (B) Probe: cDNA 3. (C) Probe: cDNA 1. (D) Probe: cDNA 4. (E) Probe: cDNA 11. (F) Probe: cDNA 15. INF, RNA from DEF cells infected with MDV. CON, RNA from uninfected DEF cells. The precise location of the probes is shown in Fig. 2. — VZ'O r .‘g-U. L.» ) £9'O '- 4C 1‘ 1'8- — 9'2 — 9'6 112- 200 112‘ 200 112- 200 112- zoo 112- 200 47 their 3’ ends (Fig. 4D). For the same reason given for the probe 1, some of the tanscript might contain more than five copies of the 132-bp direct repeat at their 3’ ends. To map 3’ ends of tanscript which terminate rightwardly within the expanded region of BamHI-H, probe 4, a 1.6-kbp AvaI-PstI fragment 3’ end labeled at it AvaI site by [a-32pijTP, was used. Twelve RNA-DNA hybrids (from 450 bases to 750 bases) were protected, indicating that the respective tanscript contained one to four copies of 132 base direct repeat at their 3’ ends (Fig. 4E). 48 FIG. 4. SI nuclease protection assay to determine the 5’ and 3’ termini of the four groups of mRNA tanscribed partly from the expanded region. (A) Schematic of the EcoRI-a subfragment of BamHI-H, with location of the probes (1 to 4) and the sites digested by S1 nuclease. The heavy line represent the expanded region, with each horizontal arrow representing one 132-hp direct repeat. Vertical arrows below each probe represent 81 nuclease digestion sites. (B, C, D, E) Result of SI nuclease assays with the probes indicated for each panel. The sizes (in nucleotides) of the protected fragment were calculated from the positions of radiolabeled DNA from a DNA sequencing reaction mm in parallel as markers (not shown). Lanes: INF, total cellular RNA from MDV-infected DEF cells; CON, total cellular RNA from uninfected DEF cells. Thirty micrograms of total RNA was used per lane. Ifi'fiu. A EcoRl l—xe Smal Aval Dral Banl Dral Pstl l—e-e-e-e-el 1 1 Baml-ll fiw—l 5’ end labeled probes 1 w* Oral-Benl 3’ end labeled probes B. Probe 1 INI= CON 5... I 400- 350— 100— C. Probe 2 D. Probe 3 INF CON INF CON _ 750 75° 730 ‘ 860 — 530 .. s40 — eso _ _ 620 - 600 600 _ 570— 580 — - 280— e- . 550- 550 - 520— 500 — 5°0_ 450 _ 475 - 450 — 27o — 150- 2 *—f—I_ DraI-Banl 3 Wat: SmaI-Banl 4 I‘m Aval-Pstl 170— o E . Probe 4 1?an CON ‘1'- 41,. 50 DISCUSSION Our results, from both cDNA analysis and SI nuclease protection studies, consistently indicate that tanscription can either be initiated or terminated within or near the expanded regions of BamHI-D and -H at multiple sites and in both rightward and leftward directions. Based on the directions of tanscription, and therefore whether they are initiated from or tanscribed toward the expanded regions, the tanscript were categorized into four groups. Figure 5 shows the precise positions of the four groups of tanscript, mapped according to the SI nuclease protection and cDNA sequence alignment. Due to the multiple tanscriptional irritations and terminations and presence of multiple BamHI-D and -H fragment containing different numbers of 132-hp direct repeat (Fig. l), RNAs with variable sizes may be tanscribed as demonstated by Northern blot analysis (Fig. 3). Previously, Bradley et al. (1) reported the result of SI nuclease protection studies indicating that a gene family, hypothesized to be directly associated with the tumorigenic potential of MDV and tanscribed to yield 1.8-kb RNA, is composed of two groups of exons that are tanscribed rightwardly from the expanded region of BamHI-H. Their observations are contadicted by the result of this report (summarized above), which were generated via cDNA sequencing combined with $1 nuclease protection analysis with conventional radiolabeled probes. Possibly the differences result from major variations in the methods used: (i) reliance on sequence analysis of cDNAs, in the study reported here, and (ii) use of SI nuclease protection studies dependent on 'bold probes" and probe generation methods that make it dificult to determine the direction or limits of tanscription, in the report of Bradley et a1. (1). Not only does our 81 nuclease protection analysis avoid such interpretation problems; this analysis is further confirmed by our use of cDNA analysis. While the 51 FIG. 5. Precise positions of the four groups of tanscript in the EcoRI-a subfragment of BamHI-H, mapped by $1 nuclease protection assay and cDNA sequence alignment with the published sequence of EcoRI-a subfragment (I), along with the assignment of the cDNAs from Fig. 2 to the four groups of tanscript, respectively. Vertical bars mark multiple initiation sites, with each bar representing one possible initiation site; arrowheads on the lines mark multiple tennination sites, with each arrowhead representing one possible termination site. Dashed lines in group 1 RNA tanscript and in cDNAs 4 and 6 are the undetermined regions. The heavy line at the top is the expanded region. The heavy arrows above the expanded region and above the cDNAs each represent one 132-bp direct repeat. Longer arrows above the map represent the direction of the polyadenylation consensus sequence. 52 n was. ..IIIIV moot. >>>4>> >>4>>> tea-... I A. _ _.+_v_v_.v.v_ _ _ _ q . 1 |J mas: ..ttlt'llullvvv mace» TTIIWTT lvl 1v 32>»... TV IV ooz> 3 To r 1. mqocou A1 1.11. ocz> #o A MIA. 9353 $1411 J. ooz> 3 m T l“ . A. A. oDz> #5 5 All I I I II II I |._ F ooz> # Allllllllalll 53 specific result reported here difier from those reported by others (1), the general suggestion that a gene family hypothesized to be directly associated with the tumorigenic potential of MDV is involved (1) deserves further consideration. Therefore, in agreement with that suggestion, and to emphasize potential significance and future research directions, the four groups of tanscript reported here will also be referred to as originating from a gene family (as described above) for discussion purposes. However, future studies will be required to identify the specific gene or genes that are actually responsible for MDV’s tumorigenic potential. In this study, a 0.67-kb RNA tanscript hybridized only with cDNA 3 (Fig. 3B), indicating that it is tanscribed on the left side of the BamHI-H expanded region without the inclusion of any 132-hp direct repeat sequence. It is possible that this 0.67-kb tanscript represent the 0.4-kb tuncated RNA tanscript of the 1.8-kb RNA family reported by Bradley et a1. (2). The size variation could be explained by the difference in size markers used. In contast to their work, this study used the RNA ladder containing markers extending below 0.67 and 0.4 kb (Fig. 3), making possible a more precise size determination. If such an interpretation is correct, the small tanscript found in this study may result from a small subpopulation of vims already having undergone genomic amplification and attenuation. This was the explanation proposed by Bradley et al. (2) for the low level of their small tanscript found in the preparations of viruses known to be pathogenic. If their hypothesis is correct, the finding of relatively minor amount of the small tanscript in this study in relation to the abundant 1.8 kb tanscript (Fig. 3B), stongly suggest that most of the vims in the preparation used was not yet attenuated. Further experiment, needed to confirm that the two small tanscript found in both studies are actually the same and to independently confirm the correlation of it presence with attenuation, are S4 beyond the scope of this preliminary study, which serves to clarify a difierent issue. Analysis of the nucleotide sequence of the 132-bp direct repeat at both rightward and leftward directions revealed several potential TATA box and polyadenylation signal consensus sequences (1 , 14; unpublished data). There is one TATA box each in both the rightward (TTATTAAAT) and leftward (TTTAATAA) directions, and two polyadenylation signal consensus sequences (15) each in the rightward (AATAAG, ATTAAA) and lefMard (CATAAA, AATAAG) directions. By correlating these consensus sequences for RNA initiation and polyadenylation with our result from $1 nuclease protection assay and cDNA analysis, it appears that these potential initiation and polyadenylation signals are likely to generate the diversity of the four groups of tanscript (Fig. 5). The concept that the 132-bp direct repeat serves as a bidirectional promoter region is not surprising, considering the observation of a similar phenomenon in the Epstein-Barr vims system (12). Their tansfection experiment, which involved the use of the chloramphenicol acetyltansferase (CAT) reporter gene, demonstated the presence of a bidirectional latent promoter region at the right- hand end of the Epstein-Barr virus genome, expressing the latent membrane protein (LMP) gene leftward and the terminal protein 2 (TP2) gene rightward (12). Since the four groups of tanscript are bidirectionally tanscribed tom the expanded regions, group 1 tanscript are complementary to group 3 tanscript and group 2 tanscript are complementary to group 4 tanscript. It has been hypothesized that the complementary tanscript, or antisense RNAs, may play a role in gene regulation. Rogers and Speck (18) reported a bidirectional tanscription of the Epstein-Barr vims major internal repeat whose rightward tanscript encode the six known viral nuclear antigens. They further 55 suggest that the leftward tanscription may antagonize the expression of those nuclear antigen messages by formation of RNA duplexes. In an experimental system with the thymidine kinase gene, thymidine kinase activity is stably reduced by the regulation of antisense RNA through formation of duplex RNA- RNA in the nucleus (10). It has also been reported (1 1) that an antisense mRNA can direct the covalent modification of the tanscript encoding fibroblast growth factor in Xenopus oocytes. It is important to note that this latter result occurs in a natural system. Although we do not know at this time whether a protein product is produced, the existence of bidirectional tanscription described in this report makes it reasonable to speculate that a similar pattern of antisense gene regulation may exist for the four groups of MDV tanscript. In summary, the work reported here, showing both rightward and leftward tanscription tom multiple sites within or near the expanded regions, raises questions concerning the result of the tanscriptional analysis by Bradley et a1. (1), in which experimental design precluded determining the direction of tanscription. Also, from our cDNA analysis and SI nuclease protection studies, it now appears that the gene family hypothesized to be directly associated with the tumorigenic potential of MDV (1) is composed of four groups of tanscript (this report) rather than two groups of exons (1). 56 IKJKNKJRHJHDGBMEIFHS Thnlremmueh “MB supponed pnnuufly by a gnuu fionrthe Emnlil Pardee Foundation, partly by grant 86-CRSR-2-2870 under the Special Research Grant program administered by the Cooperative State Research Services, Sciences and Education, of the U.S. Departnent of Agriculture, and to a limited extent by Public Health Service grant CA 45479, awarded to L. F. Velicer by the National Cancer Institute. We thank Quentin McCallum and Ruth A. Stinger for their excellent technical assistances and Peter Brunovskis, Robert F. Silva, and Richard C. Scharwtz for their careful reviews of the manuscript and many helpful suggestions. 57 Bruiley.G..Mnyashi,G.Lancz.A.Tanaka.andM.Nonoyana. 1989. Stucture of the Marek's disease virus BamHI-H gene family: genes of putative importance for tumor induction. I. Virol. 63:2534-2542. Bradley, G.,G.Lancl,A.Tanaka.andM.Nonoyama. 1989. Loss of Marek's disease virus tumorigenicity is associated with tuncation of RNAs tanscribed within BamHI-H. I. Virol. 83:4129-4135. B. W. Calnek 1985. Marek’s disease - a model for herpesvirus oncology. Crit. Rev. Microbiol. 12:293-320. Glutcltill, A. B., R. C. Chubb. and W. Baxendale. 1969. The activation, with loss of antigenicity of the herpes-type virus of Marek’s disease (stain I-IPRS-l6) on passage in cell culture. I. Gen. Virol. 4:557-564. Fuhrchi. I., M. Sudo, Y.-B. Lee, A. Tanaka.and M. Nonoyama.1984. Stucture of Marek’s disease virus DNA: Detailed restiction enzyme map. I. Virol. 81:102-109. WK..ATu1aka.L.W.SchiemIan.R.L.Witter,andM.Nonoyama. 1985. The structure of Marek’s disease virus DNA: the presence of unique expansion in nonpathogenic viral DNA. Proc. Natl. Acad. Sci. USA 82:751- 754. Glaublger, C., K. Naserlan. and L. F. Velicer. 1983. Marek’s disease herpesvirus. IV. Molecular characterization of Marek's disease herpesvirus A antigen. I. Virol. 45:1228-1234. Isfort. R. I., H.-I. Eng, and L. F. Velicer. 1987. Identification of the gene encoding Marek's disease herpesvinls A antigen. J. Virol. 61 :2614-2620. Kato. S.. and K. Hirai. 1985. Marek’s disease vinls. Adv. Virus Res. 30:225- 277. 10. 11. 12. 13. 14. 18. 18. 17. 18. 19. 58 En. S. E. and B. I. Wold. 1985. Stable reduction of thymidine kinase activity in cells expressing high levels of anti-sense RNA. Cell 42:129-138. Emelman, D., and M. W. Enchner. 1989. An antisense mRNA direct the covalent modification of the tanscript encoding fibroblast growth factor in Xenopus oocytes. Cell 89:687-696. Law. 6., A. Eoonomou. and P. I. Farrell. 1989. The terminal protein gene 2 of Epstein-Barr vims is tanscribed from a bidirectional latent promoter region. I. Gen. Virol. 70:3079-3084. Maruatis, T., E. F. Hitch. and I. Sambroolt. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. Maotani.K.,K.Kanamorl.K.nnlta,S.Ueda.S.Kato,endK.Hira1.1986. Amplification of a tandem direct repeat within inverted repeat of Marek’s disease virus DNA during serial in vito passage. I. Virol. 88:657-659. Montell, C., E. F. Halter, M. H. Caruthers, and A. I. Berk. 1983. Inhibition of RNA cleavage but not polyadenylation by a point mutation in mRNA 3’ consensus sequence AAUAAA. Science 308:600-605. Nasarlan. K. 1970. Attenuation of Marek’s disease virus and study of its properties in two different cell cultures. I. Natl. Cancer Inst. 44:1257-1267. Okazaki. W., W. G. Purchase. and B. R. Burmesier. 1970. Protection against Marek’s disease by vaccination with a herpesvirus of turkeys (HVT). Avian Dis. 18:108-125. Rogers. R. P., and S. H. Speck. 1990. Bidirectional tanscription of the Epstein-Barr virus major internal repeat. I. Virol. 64:2426-2429. Sanger, F., S. Nlcltlen, and A. R. Coulson. 1977. DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. 21. 59 Shanna. I. M., and H. A. Stone. 1972. Genetic resistance to Marek’s disease. Delineation of the response of genetically resistant chickens to Marek’s disease virus infection. Avian Dis. 18:894-906. Silva. R. F., and R. L. Witter. 1985. Genomic expansion of Marek’s disease virus DNA is associated with serial in vito passage. I. Virol. 84:690-696. Stuart. P., M. Ito, C. Stewart, and S. E. Conrad. 1985. Induction of cellular thymidine kinase occurs at the mRNA level. Mol. Cell. Biol. 8:1490-1497. Chapterm IdenfificationofaUniqueMarek'sDieeaeeVirueGener‘lichEncodesam ElodaltonlesphoprotinuldhExpreesednbofltLyflcallyhlfectheflsand LatentlylnfectedLymphoblastoidTlnnorGells Xinbin Chen, Paul I. A Sondermeijer, and Leland F. Velicer. 1992. I. Vlrol. 66:85-94. 60 61 Dr. P. I. A. Sondermeijer provided 2.6-kbp nucleotide sequence of BamHI- H and helped writing this manuscript. 62 ABSTRACT The Identification of unique Marek’s disease (MD) vims (MDV) antigens expressed not only in lytically infected cells but also in latently infected MD lymphoblastoid tumor cell lines is important in understanding the molecular mechanisms of latency and tansformation by MDV, an oncogenic lyrnphotopic herpesvirus of chickens. Through cDNA and nucleotide sequence analysis, an open reading frame (designated the pp38 ORF) which encodes a predicted polypeptide of 290 amino acids was identified in BamHI-H. Demonstation that the pp38 ORF spans the junction of the long MDV unique and long internal repeat regions (MDV has an alphaherpesvirus genome structure) precludes the presence of the gene encoding the B antigen complex (gp100, gp60, and gp49) in the same region of BamHI-H, where it was originally thought to exist. Duplication of the complete pp38 ORF was not observed in BamHI-D, but part of it (encoding 45 amino acids) was found in the long terminal repeat region of the fragment. By use of trpE-pp38 fusion proteins, antisera against pp38 were prepared. By irnmunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electophoresis, a predominant virus-specific 38,000 Da polypeptide (designated pp38) and a minor 24,000 Da polypeptide (designated p24), were found. No precursor-product relationship was found between pp38 and p24 by pulse-chase analysis, and only pp38 was detected by Western blot analysis with antiserum to pp38. pp38 was found to be phosphorylated and present in oncogenic serotype-l but not in nononcogenic serotype-3, MDV-infected cells. Expression of the gene encoding pp38 was relatively insensitive to phosphonoacetic acid inhibition, suggesting that pp38 may belong to one of the early classes of herpesvirus proteins. pp38 was also detected in the latently infected MSB-l lymphoblastoid tumor cell line. The detection of antibody against pp38 in immune chicken sera indicates that pp38 is an immunogen in 63 birds with MD. Most of the properties described here for a protein detected by methods based on finding the ORF first are identical to those of a 38 kDa phosphoprotein reported by others, suggesting that they are the same. Collectively, the data reported here provides (i) more definitive information on the complete ORF of another MDV gene and the protein that it encodes, (ii) clarification of the gene content within a specific region of the MDV genome, and (iii) the molecular means to conduct further studies to determine whether pp38 plays a role in MDV latency and tansformation. 64 INTRODUCTION Marek’s disease (NID) is a lymphoproliferative disease of chickens caused by cell-associated MD vims (MDV) and is characterized principally by T-cell lymphomas and peripheral nerve demyelination (4). The T-cell lymphomas occurs after the establishment of latency in the MDV-infected chickens, raising questions regarding the viral gene(s) expressed to cause these two phenomena. Presumably one or more viral genes are responsible for the establishment and maintenance of latency and for tumor induction. Whether these event are the result of the same gene or separate ones is unknown. Identification of an MDV-specific antigen(s) expressed not only in MDV-infected fibroblast cells but also in latently infected and tansformed MD lymphoblastoid tumor cell lines and MD tumors is an important first step toward an understanding of the mechanisms of latency and tansfonnation by MDV. In lytically infected cells, there is extensive gene expression over the entire MDV genome (19, 29). In latenfly infected lymphoblastoid tumor cell lines derived from MD tumors, 4 to 7 tanscript were reported by Schat et a1. (29) in their partial tanscriptional mapping of MDV, 32 tanscripts clustered in the short and long repeat regions were reported by Sugaya et al. (33), and 29 tanscript spread over almost the entire MDV genome were reported by Maray et al. (19). Very few of the tanscript detectable in both lytically and latently infected cells were further characterized. Relevent to this study was the report of a 1.9-kb immediate early (IE) tanscript localized to BamI-II-H (29). Silva and Lee (31) first reported the existence of three viral proteins of 41, 38, and 24 kDa ( then designated p41, p38, and p24, respectively) detected in serotype-l MDV (oncogenic MDV)-infected cells but not in fibroblast cells infected with nononcogenic serotype 2 MDV and nononcogenic serotype 3 herpesvims of turkeys (HVT). p38 was the most prominant molecule. Later, 65 Ikuta et al. (13) and Nakajirna et al. (22) reported a similar group of MDV specific antigens of 43, 39, 36, and 24 kDa also detected in latently infected MSB-l lymphoblastoid tumor cells, but only when the cells were teated with 5- iodo-2-deoxyuridine (IUdR). The latter workers reported that these proteins were phosphorylated at serine residues. Furthermore, on the basis of in vitro tanslation analysis, they suggested that the 24-kDa and 39- or 36-kDa (their newer terminology) phosphoproteins were tanslated from distinct mRNAs and encoded from overlapping genes or from separate regions with partial DNA homology within the MDV genome (22). The MDV-specific phosphoproteins were also detected in tumor lesions of chickens with MD (21). Cui et al. (6) further reported the identification of two Agtll clones which may encode p41, p38, and p24 (31). Both clones were mapped to the BamHI-H fragment of the MDV genome. Five proteins (135, 41, 38, 24, and 20 kDa), including p38, were detected by antisera against two MDV-lacZ fusion proteins (6). However, the nucleotide sequences of the two clones were not determined, the structural relationship among the five proteins, if any, remained to be elucidated, and the entire gene (8) encoding the phosphoprotein(s) was not defined (6). In this study, an open reading frame (ORF) was identified in the BamHI-H fragment of the MDV genome through cDNA and nucleotide sequence analyses. The predicted protein product of the ORF was detected in both lytically and latently infected cells and was found to be 38 kDa, phosphorylated, not detected in HVT-infected cells, and immunogenic in birds with MD. 66 MATERIALS AND METHODS Cells and viruses. The preparation, propagation, and infection of duck embryo fibroblast (DEF) cells with MDV and HVT were performed as described previously (5). The MSB-l cell line (1) was used as a representative MD lymphoma-derived cell line and was cultured in RPMI 1640 supplemented with 10% fetal calf serum at 370C in a humidified atnosphere of 5% C02 in air. At the time at which the MSB-l cell line was used, it infection with MDV was latent, as shown by a lack of the MDV A antigen (gp57-65), the MDV B antigen complex (gp100, gp60, and gp49) and MDV p79. The MDV GA stain used in this study was at cell culture passage 6 following the isolation of cell-free virus from feather tips obtained from infected birds showing symptoms of MD. The HVT FC-126 vaccine stain vims at cell culture passage 13 was used in this study. Antisera. The immune chicken sera (ICS) were convalescent-phase sera obtained from chickens with MD as a result of natural infection. ICS contain antibodies against the MDV A antigen (gp57-65), the MDV B antigen complex (gp100, gp60, and gp49), MDV p79 and other MDV proteins that remain to be characterized. Construction and saeening of the cDNA library. A cDNA synthesis kit from Bethesda Research Laboratories (BRL) was used in accordance with the manufacturer’s specifications. The newly synthesized cDNAs were made blunt- ended via the fill-in reaction catalyzed by the Klenow tagment of Escherichia coli polymerase I. The blunt-ended cDNAs were directly cloned into the SmaI site of calf intestine alkaline phosphatase-teated pUC18. An in situ DNA hybridization method was used to screen the cDNA library and was carried out essentially as described previously (5). 67 DNA seqileridng and nucleotide sequence analysis. The initial cDNA sequencing was carried out with the pUC sequencing system as previously described (5). Sequencing of genomic DNA and further sequencing of cDNA were carried out with the M13 sequencing system as previously reported (28). The BamHI-H fragment used in this study is from the BamI-II library of the MDV GA stain (kindly provided by M. Nonoyama). Genepro software (Riverside Scientific Enterprises) was used to perform the ORF analysis. For protein homology analysis, a search of the National Biochemical Research Foundation- Protein database (Release 21.0, 6/89) was made with the Genetics Computer Group program FASTA from University of Wisconsin. RNA isolation and Northern blot analysis. Total cellular RNA was isolated from mock- and MDV-infected DEF cells essentially as described previously (5). Purification of poly(A) + RNA was carried out as described previously (27). RNA was size fractionated on a 1.2% agarose gel containing 2.2 M formaldehyde and was tansferred to a nitocellulose filter. The Northern blot was hybridized (probes are described in Figure legends) and washed as described previously (5). Transcript size determinations were based on a comparison with a BRL RNA ladder run in parallel. Sl nuclease analysis. S1 nuclease analysis was carried out essentially as described previously (27) with the probes described in the text. The radiolabeled DNAs from a DNA sequencing reaction were used as size markers. Hybridizations were performed at 46°C, a temperature chosen because of the G+ C content of the MDV DNA probes. Expredon of tpE fusion protins. The vector system used to express the MDV ORF in E. coli consist of a group of plasmids (pATH vectors) encoding approximately 37 kDa of the bacterial tpE ORF under the contol of the inducible tp operon promoter (17). A polylinker with multiple cloning sites 68 at the 3’ end of the tpE ORF allows iri-frame insertion of a foreign ORF. Five DNA fragment from the MDV ORF were inserted into various tpE pATH vectors, and their respective tpE fusion proteins were generated. The precise locations of these DNA fragment and the sizes of their respective tpE fusion proteins are detailed in Result. Antibody [seduction Fusion proteins were purified by electophoresis in preparative 7.5% sodium dodecyl sulfate (SDS)-polyacrylarnide gels. The gels were stained with Coomassie brilliant blue to identify the positions of the fusion proteins. The areas of the gels containing the fusion proteins were excised and emulsified in complete Freund’s adjuvant (Difco Laboratories). Initial immunization of 2-kg New Zealand White female rabbit was done with 200 to 500 pg of protein. Rabbit were boosted after 3 weeks with 100 to 200 pg of protein in incomplete Freund’s adjuvant, and sera were collected 7 to 10 days later. Subsequent booster immunizations were done approximately 2 weeks apart as needed to obtain detectable and usable titers. Radiolabeling of proteins. Mock- and MDV-infected DEF cells were labeled with [35$]methionine (specific activity, 1,000 pCi/mmol, ICN) at 48 h postinfection for 4 h by a standard method as previously described (10). MSB-l cells (3 x 106 cells per 60-min plate) were labeled with 100 pCi of [35$]methionine per plate for 4 h. Similar method was also used for 32P- labeling of DEF cells, with following exceptions. Phosphate-free Dulbecco modified Eagle medium was used for l h before and during the 4 h labeling and 500 p61 of 32Pi (carrier-free, Amersham Corp.) was used per 60-min plate. Pulse-chase labeling was done by method already established for the MDV system (10), with following minor modifications. Preincubation without methionine was for 1 h, pulse-labeling was done by incubation with 250 pCi of [35$]methionine in 1 ml of medium per 60-mm plate for 5 or 15 min, and chases 69 were done in the presence of unlabeled methionine for 0, 5, 15 or 30 min. For labeling in the presence of phosphonoacetic acid (PAA) (Sigma) at concentation of 200 pg per ml (23) or 2 pg of tunicamycin (Calbiochem—Behring Corp.) per ml, cells were preincubated with PAA for 24 h or with tunicamycin for l h and teated at same concentation during the 4-h labeling period. After labeling was complete, the culture medium was collected, cells were washed three times with sterile phosphate-bufl'ered saline, and ice-cold detergent buffer was added to lyse the cells (10). The culture medium samples and lysates were clarified by centifugation as previously reported (10). Imminoprecipitaiion and SDS-PAGE analyses. Irnmunoprecipitation analysis was carried out as previously described (10). The irnmunoprecipitates were washed, suspended in sample buffer, and analyzed by SDS-polyacrylamide gel electophoresis (PAGE). Protein markers (BRL) were used as molecular weight standards. Molecular sizes were calculated by interpolation between standard proteins by the method of Weber and Osborn (34). Fluorography was carried out as previously described (2). 10% SDS-polyacrylamide gels were used unless indicated otherwise. Anti-pp38, an antiserum against one of the tpE fusion proteins of the MDV ORF was used unless indicated otherwise. Western blot analysis. Western blot analysis was carried out essentially as described previously (27). ass-labeled infected cell lysates were used for irnmunoprecipitation analysis with anti-pp38, and the immunoprecipitates were run on a 10% SDS-polyacrylamide gel. Proteins were blotted onto a nitocellulose filter. The filter was probed with anti-pp38 at a 1:400 dilution. Horseradish peroxidase conjugated goat anti-rabbit immunoglobulin G (Sigma) was used as the second antibody. Bound second antibody was detected by subsequent incubation in the presence of the substate solution (4-chloro-l - naphthol) (Sigma). 70 Nucleotide sequence accession number. The nucleotide sequence reported in this article has been given GenBank accession number M73484. 71 .ULTS IsolationandlocalisatlonochNA2anchNA7. TheMDV alphaherpesvinls genome stucture and the locations of BamHI-D and -H within the MDV genome (3) are presented in Fig. 1A. The 917-bp EcoRI c subfragment of BamHI-H (Fig. 1B) was used to probe the MDV cDNA library in pUC18. Two cDNAs were identified by in situ hybridization and designated cDNA 7 (297-bp) (Fig. 1C) and cDNA 2 (1,092-bp) (Fig. 1D). Both cDNAs were completely sequenced. Alignment of the two cDNA sequences with the sequence of BamHI-H (see below) revealed the precise locations of the two cDNAs (Fig. 1C and 1D). The 3’ end of cDNA 2, presumably the 3’ end of its mRNA. was located in the EcoRI b subfragment 567 bases to the left of an EcoRI restiction site designated EcoRI site-2, which separates the EcoRI b and c subfragment of BamHI-H. The sequence alignment also revealed that cDNA 7 is a part of cDNA 2 (Fig. 1C and 1D), and that the direction of tanscription for their mRNA is from right to left in relation to the MDV genome (Fig. 1E). Detection of the tramcript for cDNA 2. Since cDNA 7 is a part of cDNA 2, only the latter was used as a probe for Northern blot analysis. A major 1.95- kb tanscript was detected along with two minor component at 1.3- and 0.8-kb (Fig. 2A, data from a long exposure are presented to demonstate the existence of the latter two). When the EcoRI c subfragment was used as a probe (Fig. 1B), only the 1.95-kb mRNA was detected (Fig. 2B), even upon prolonged exposure (data not shown). Since 525-bp of cDNA 2 is located in the EcoRI c subfragment (Fig. 1D) and only the 1.95-kb mRNA was detected by both probes (cDNA 2 and EcoRI c subfragment), the 1.95-kb major tanscript (Fig. IE) is the mRNA from which cDNA 2 is generated. DeflningtheB’endofthe 1.98-kmeNAbySl nuclease analysis. On the basis of the location of 1,092-bp cDNA 2 (Fig. 1D) and the size (1.95-kb) of the 72 FIG. 1. Organizational summary of the MDV alphaherpesvinis genome stucture (3), restriction enzyme map of BamHI-H, locations of the two cDNAs, and locations of the mRNA and coding region for the pp38 gene. (A) Schematic representation of the MDV genome. The MDV genome consist of a UL region, flanked by a TRL and an IRL, and a short unique region (Us), flanked by a short inverted repeat (le) and a short terminal repeat ('I‘Rs). D and H represent the locations of the BarnI-II-D and -H fragment in the MDV genome, respectively. (B) Restiction enzyme map of BamHI-H. The EcoRI a, b, and c subfragment of BamHI-H are labeled as such. (C) Location of cDNA 7. (D) Location of cDNA 2. (E) Location and direction of tanscription of the 1.95-kb mRNA. (F) Location of the pp38 ORF. 73 1. Fig. : 0 Cr. _ — Fhm S _. 5 _. _ a m «a m mates 9:17: c r 5 _. _ . _ _ .éihfi min .éihn moat mafia 021: main E». U 231M moon. 0 93¢ mafia m 8% V Moves 82) m $8-8 comm auto amz.» m. 1| w. some can muses 74 FIG. 2. Northern blot hybridization for the detection of the tanscript of cDNA 2. (A) cDNA 2 was used as the probe. (B) EcoRI c subfragment of BamHI-H was used as the probe. A total of 0.5 pg of poly(A)+ RNA was used per lane. INF, RNA from MDV-infected cells. CON, RNA from mock-infected cells. (base) Em CON PROBE Fig. -950 -370 76 tanscript, the 5’ end of the mRNA was predicted to be located adjacent to an EcoRI restriction enzyme site, designated EcoRI site-1, which separates the EcoRI c and a subfragment of BamHI-H (Fig. 1B). Two difierent probes were used to determine more precisely the 5’ end of the 1.95-kb tanscript. A 1.87-kbp EcoRI-Smal subfragment resulting from partial digestion of BamHI-H (Fig. 1B) was 5’ end labeled. Following it digestion by EcoRI, a 917- bp EcoRI-EcoRI (EcoRI c) subfragment was 5’ end labeled only at EcoRI site-2. The result of an 81 nuclease assay indicated that this 917-bp EcoRI c subfragment was fully protected (data not shown), suggesting that the 5’ end of the 1.95-kb tanscript is located to the right of EcoRI site-l in the EcoRI a subfragment. To further map the 5’ end, an EcoRI a subfragment was 5’ end labeled. Following it digestion by Smal, a 950-bp EcoRI-SmaI subfragment (Fig. 1B) was 5’ end labeled only at EcoRI site-l. A 370-bp RNA-DNA hybrid was protected from $1 nuclease digestion (Fig. 3), indicating that the 5’ end of the 1.95-kb mRNA is located in the EcoRI a subfi'agment 370 bases to the right of the EcoRI site-l. These result indicate that the tanscript extends from its 5’ end, 370-bp to the right of the EcoRI site-l, through the 917-bp EcoRI c subfragment, as determined both by $1 analysis, to its 3’ end, 567-hp to the left of the EcoRI site-2 as determined by cDNA analysis. Therefore the size of the mRNA is 1.85-kb, not including the poly(A) tail, which presumably contibutes to the 1.95-kb of the actual tanscript. Nucleotide sequencing and ORF analysis. Both cDNA 2 and its corresponding region on BamHI-H were completely sequenced in both directions. After sequence alignment, only 2 nucleotides were found to differ between cDNA 2 and the genomic DNA (Fig. 4): a C in cDNA 2 compared with an A in the genomic DNA at position 1,045, the third position of a codon within a predicted ORF, resulting in no predicted amino acid change; and a T in cDNA 77 FIG. 3. SI nuclease protection assay for the determination of the 5’ end of the 1.95-kb mRNA. A 950-bp EcoRI-Smal subfragment of BamHI-H (Fig. 1B) was used as the probe. The sizes (in nucleotides) of the protected fragment was calculated from the positions of radiolabeled DNAs from a DNA sequencing reaction run in parallel as a marker (not shown). Lanes: PROBE, undigested probe contol; CON, total cellular RNA from uninfected DEF cells; NF, total cellular RNA from MDV-infected DEF cells. Thirty micrograms total RNA was used per lane. Fig.“ A" Probe: Size (kb) I N F cDNA 2. C O N Probe: EcoRI c Subfragment I c N o F N 79 FIG. 4. Nucleotide sequence of the pp38 gene and analysis of it ORF. The 5’ end of the mRNA determined by $1 nuclease analysis is numbered as nucleotide l. The predicted amino acid sequence of the ORF is shown by the single-letter code above the nucleotides with amino acid numbers in parentheses. The potential TATA box and AATAAA consensus sequence for polyadenylation are underlined. The two EcoRI restriction enzyme sites are underlined and labeled as such. The vertical bar followed by "Beginning of cDNA 2" represent the start of cDNA 2. The two nucleotide differences found between genomic DNA and cDNA 2 are underlined at nucleotide positions 1,045 and 1,516. EN) 1114;. 41. ~800 -700 -600 -500 -400 -300 -100 101 201 $11 131’ 131’ 33?) 53121 ,3... 611” silt) iiS?’ 1531’ 1301 1401 1501 1601 1701 1801 ACAIAITTITCCATGTAAICAACATTCGCAGAATAAACCTTCCATTTTAATGATCGCGGTCCTATATTGTGAACTGICCCCCAACAAAAAAAAAIATCSS AITATAICAGCCCATCCTTTCTACATTGCACGACCCAGCGCGTCGCTCATTCCIECGATAAAAGACCATAACAIGABCAAATGAGACCAIACAGAAACGA CACCCATGGGTCTGCCCAGCAGGCTGAYCCCGGGTCGATGTTGACGGTGGTCTGCGGTCCCGGATCGTCGGGACTGACCGACGAACCGCTICAICATCAA ATATCGCCGTAAAIAAAAAATCGCCGTTCIATIAATIGAGATCCTTTTTTTTATTTGCTTATCGAATGCTAGTAAITITATTACTTAITTGATGAAGGGA GAAATTCASCTCGTTCGTCGCCTAGCGTACGTTCCIIACAGGAAATATATCGGGGATCGGCCGTGCCATTCIGAGAGAGCATCGCGAAGAGAGAAGGAAC CTCGCAACCGCCGCTCTTTTATACACAAGAGCCGAGCCGCCCCCACATGTACCCCCAACACICAAGTGCGAATTTGGGGCGGTACAIGTCACGTGATAAC ATATCGCCATATCCGATTGGCTCACCICGGCGIICGCACCAGAGTCCAATAATATAAIATAATATAATATATTATTGGTTCGCAGTGCGAACGCTGACGC GTTCGCACTGCTCAIITGCATACACAICACGTGATAGTICGAGTAGGCGGIACGCCCACCCGIAIAASAAICGTAATTTCTTGTGGCCICGAGTGGCGGT SCGACTTGCTCICGICGGACGGGAGGCGGCGGTAIAGGAIAAGAGATCACAAAAAAGCGAGACCTGGAICGAACGGCAACGTCTCGTCCCGGITGTTAAT CSTTIGGAACTTCTTCGCCIGATCGGTGGIGTAACCGIGTAGGTATITTIAGTTITTAIGTACCATTICGGTTSCITTAITATATIICCCACCCAICGTI TITCTTIAIAACAIAGITTGACCCTCTCGGCATCACAGATGSCACCCTCCCTCCACTCATGACCCACAACCGCGATTTGTTTTTCATCTTCAACCCACAG CCAICCTIGTCTITCTGCCCGCACCGCACGCTTTGCTCGTCCCCGCGTGCAAGAICGGCAGGGGGTGATGEQGEIQEAASCAEAAEACEAASGGETGICG retain.casement.antennaesmmechntahs.enumerates eWiessscssrdeeetminassessments:misintescciehsasex. assessed.rsmiths.i..i..i..t..l..2..iee..i..h.l..iai..ioi..i..bassassins. assessments"thist.rhas.sentecnnmeggegggggig,twininmmsi GIAIGEAGEAGAIGEGGEAGLTISCCEAGEAGIGCEAASGAEGAACAIATeCGgACIIGETTXTCEAASCAEAGEAASCTXITXTAEATICCXTTEGCSC AITAAIGtTGSCCEAAAGAEAAAACECAAAIATAITGEGGEAGEATITGAAIAAAAAAEGGXIISTIXIAEAASGAECCEGTICTAIIETAICCXIGEAG iciitcitttctlclitcictirrittiteitcittlclitoiléietictécritticlitrittttlicricciecitcictitritcéctfiotitci IAXT0267AGAICGSCASACXTAAAAICAECAITAYGGZATICIXTAIGTITGITAATGSCTITCIGISCAEGCAITXTCXTTSGGSGAXTGBAIICIEG GEAGXTGEAAICTEGAEAAICAAAA7CTEAAICAAATIAAAITTAATACAGTGTAGCCGTACCCGACGTTGGAGGCGGAGATIAAGCEéQIIQICACCTT IACGAATATIGSTGCAGACAAAGACCAAAAAATGGAAAATGGACAGCTGCAGCACGAAAGTCTCGATTTGGATGCAGATGCCGTTTCTATACCCGAGACI ATCTCCCCACCAATCGAGGAAGAACCIGIGCTIICAGATATTGATGAACAATCAGAAIATATICAITIACAATTAGAATCGGTTACCAGATACAATAATT CCBCACTGIIGCCCA$ATACGATGATGCAGTTGACCCACCCCCTTCAIACGATTCCCTATCCCCGATACATAAIGTTAACAATTCTGAAAGTTGCGCAGA ABTTGACTIGCGTTTYATCATTCGACATGATGGAIGTGCGATCGCTACATIATTAATACTTTTTTTGACGGTAGTTTCTGCAACCCTIGTAACTATTAIC ACAGAAACATAAITGACGTATGTGATACAATAAAIATGCACGTCTGATCCCAAITGAGACTTTTATGTTCIATGACGAIACTAACAAGGTGTAGGTTITA casrrrccrasrrrsrAcrccrAAiccxrarrccagglgggrnsrrrcrrrrsr 81 2 compared with a C in the genomic DNA at position 1,516, which is in the 3’ noncoding region of a predicted ORF. Analysis of the DNA sequences of cDNA 2 and BamHI-H (Fig. 4) resulted in identification of an ORF, designated pp38 ORF (Fig. 1F), which encodes a predicted polypeptide of 290 amino acids. The first in frame ATG initiation codon of the pp38 ORF is located just 3 nucleotides upstearn of the EcoRI site- 1 (Fig. 4). A 367 base 5’ untanslated leader sequence and a 617 base 3’ untanslated sequence were found in the 1.95-kb mRNA (Fig. 4). A potential TATA consensus sequence is located 38-bp upsteam of the mRNA initiation site (Fig. 4). An AATAAA motif for polyadenylation was found at position 1,836, 13- bp before the start of the poly(A) tail in cDNA 2 (Fig. 4); this position is within the general range of 10 to 30 bases upstearn of the 3’ end before the addition of poly(A) tail (20). The calculated molecular weight of the predicted 290 amino acid polypeptide is 31,169. Only one potential N-linked glycosylation site was found at predicted amino acid position 214 (Fig. 4). Upon hydropathy analysis, the carboxy terminal end of the predicted polypeptide was found to contain two hydrophobic potential tansmembrane domains (data not shown). No signal peptide was found. As determined by GenBank protein homology analysis, the predicted 290 amino acid sequence lacks any significant homology to any other known protein sequences, including those of other herpesviruses and oncoproteins. Generationoffiisionproteinsandproductlonofantibodies againstthem. The locations of the potential epitopes predicted from the antigenicity plot are presented in Fig. 5A. Five DNA fragment (Fig. 5B, fragment 1, 2, 3, 4, and 5) from the pp38 ORF coding region were cloned into various pATH vectors, and five respective trpE fusion proteins were generated (Fig. 5C, lanes 1, 2, 3, 4 and 5). Because of the initial low expression of the tagment 3 fusion protein (Fig. 82 FIG. 8. Antigenicity profile of the pp38 ORF and generation of tpE fusion proteins. (A) Antigenicity plot of the predicted amino acid sequence encoded by the ORF. The antigenicity plot was obtained with the Hopp-Woods program (12). (B) Positions and sizes of DNA fragment 1, 2, 3, 4, and 5 used for the generation of tpE fusion proteins. (C) SDS-PAGE analysis of fusion proteins on a 7.5% SDS-polyacrylamide gel. Lanes: C, bacterial stain RRl protein; T, 37 kDa tpE protein; 1, 2, 3, 4, and 5, tpE fusion proteins of fragment 1, 2, 3, 4, and 5, respectively; 3A. an additional tpE fusion protein of fragment 3. Fusion proteins are indicated by dot. The gel was stained with Coomassie brilliant blue. (.1 Ln tn :oded Fig. 5 an A. The Hopp and Woods Antigenicity Profile (the 2" aim rm: 1.5 sm 1.. 137 l 5 H, s.e -D.5 m -I.O 1m —t.s se ‘ Billing?” m 5. B. The Locations and Sizes of DNA Fragments Used to Generate TrpE Fusion Proteins from amino acid 2 to 62 from amino acid 65 to 114 from amino acid 116 to 182 from amino acid 133 to 290 from amino acid 2 to 290 UIIhUNt-l essa- C. SDS—PAGE Analysis of TrpE Fusion Proteins CT1233A45M(kd) —97.4 -68 -37 84 5C, lane 3), an additional fusion protein of fragment 3 was produced (Fig. 5C, lane 3A). Antibodies against these six fusion proteins were produced in rabbits as described in Materials and Methods, and the sera were designated antisera 1, 2, 3, 3A. 4, and 5, respectively. Identificationofpp38astheorilygeneproductofthepp380RF. Irnmunoprecipitation analysis of culture media and lysates from mock- and MDV-infected DEF cells was done with all six antisera produced against the pp38 ORF tpE fusion proteins. No MDV-specific protein was irnmunoprecipitated from the culture media by the six antisera (data not shown). However, one predominant 38 kDa protein (designated pp38 on the basis of phosphorylation analysis, see below) and a minor 24 kDa protein (designated p24) were found in irnmunoprecipitates from the MDV-infected cell lysates by five of the six antisera (Fig. 6, lanes 1, 3, 3A. 4, and 5). The antisenlm against the fragment 2 fusion protein (Fig. 6, lane 2) was negative for these two proteins, supporting the prediction by the Hopp-Woods program (12) that a major portion of fragment 2 is poorly antigenic (Fig. 5A). Antibody titation analysis indicated that antiserum 5 contained the highest antibody titer among the five positive antisera (data not shown). Therefore, only antiserum 5, designated anti-pp38 hereafter, was used in the subsequent irnrnunoprecipitations. Pulse-chase labeling analysis was performed to determine whether there was any precursor-product relationship between pp38 and p24. Both mock- and MDV-infected DEF cells were pulse labeled for 15 min and chased for 0, 5, and 15 min. Lysates from labeled cells were subjected to irnmunoprecipitation analysis with anti-pp38 (Fig. 7A). There seemed to be no apparent relationship between pp38 and p24. pp38 was partly degraded during the first 5 min of the chase and then remained stable. However, no concomitant increase in p24 was 85 FIG. 8. Identification of pp38 as the protein product of the pp38 ORF. Lanes: C, lysate from mock-infected cells subjected to irnmunoprecipitation with antiserum 5 (anti-pp38); l to 5, lysates from MDV-infected cells subjected to irnmunoprecipitation with antisera 1 to 5, respectively. pp38 and p24 represent 38- and 24-kDa proteins found in irnmunoprecipitates with antisera l, 3, 3A. 4, and 5, respectively. .1 . 21g. C. C1233A48 913; HW o '4 ts " ' 43- 18.4- 87 FIG. 7. Identification of pp38 as the only gene product of the pp38 ORF (A) Kinetics of MDV pp38 processing, as determined by pulse-chase analysis. A 15 min pulse was followed by chases of various times shown above each lane. Contols were labeled for 4 h. NF, MDV-infected cell lysates. Lanes: -, uninfected; +, infected. (B) Western blot analysis of pp38 with anti-pp38. Lanes: NF, anti-pp38 irnmunoprecipitate from an MDV-infected cell lysate; CON, anti-pp38 irnmunoprecipitate from a mock-infected cell lysate. (C) Autoradiographic analysis of the filter used in the Western blot analysis. p79, pp38, and p24 represent 79-, 38-, and 24-kDa proteins, respectively. 7. P 5:38 on uewm 3008an w. 2369 Box nbfiwcmwm O. Wfionmanonnmeao mafia om ptwm 0». 9o Hum—8n Emma F 3389 92 ahead? fimsmm H N w A m m u m H n H n wchmABHDV 5 t 3 z o z o 03mmm2=55 o o m m Hm Hm Hmtmw m. z m. z sz I + I + I + I + pqmi III scum Moms: - I 8 8 punt .I.. rig. 89 observed. Similar result were obtained when 5 min of pulse labeling along with chases for 0, 5, 15, and 30 min were used (data not shown). Western blot analysis was conducted to determine whether both pp38 and p24 are protein product of the pp38 ORF. Only pp38 was observed in MDV-infected cell lysates when the nitocellulose filter was probed with anti- pp38 (Fig. 7B, lane NF). As determined by autoradiographic analysis of the filter used for the Western blot, the amount of p24 tansferred to the filter was quite substantive (Fig. 7C). Note also that, p79, a MDV-specific protein that always tends to be nonspecifically tapped during irnmunoprecipitation (Fig. 7C) served as an internal negative contol. Despite the abundance of both p24 and p79 on the filter, neither was detected by anti-pp38 on the Western blot. Characterisation of pp38. To determine whether the 38-kDa protein identified in this study is phosphorylated, we irnmunoprecipitated lysates from 32P-labeled mock- and MDV-infected DEF cells with anti-pp38 (Fig. 8A, lanes 1 and 2). Lysates from 35S-labeled mock- and MDV-infected DEF cells were used as contols (Fig. 8A. lanes 3 and 4). The result of SDS-PAGE analysis showed that pp38 is phosphorylated (Fig. 8A, lane 2). An additional phosphorylated protein was found at the 31 kDa position in infected cells (Fig. 8A. lane 2), but it seems to be a cell protein since a similar protein was slightly labeled in the contol lane (Fig. 8A. lane 1). There was little, if any, 3ZP-labeled protein at the 24 kDa position in either contol or infected cell lysates (Fig. 8A. lanes 1 and 2). Generally, PAA-insensitive herpesvirus genes tend to encode one of the early classes (IE or early) of herpesvinls proteins (9). To determine whether the expression of the gene encoding pp38 is sensitive to PAA teatrnent, we irnmunoprecipitated lysates from teated and unteated cells with anti-pp38 (Fig. 8B, lanes 5, 6, 7, and 8). As positive late protein contols (Fig. 8B, lanes 1, 2, 3, 90 FIG. 8. Characterization of pp38. (A) pp38 is a phosphoprotein. Lanes: 1 and 2, lysates from 3zP-labeled mock- or MDV-infected cells, respectively; 3 and 4, lysates from 35S-labeled mock- or MDV-infected cells, respectively. (B) Efiect of PAA on the synthesis of pp38. Lanes: 1 and 2, culture media from 358- labeled, PAA-teated, mock- or MDV-infected cells; 3 and 4, culture media from 35S-labeled, unteated, mock- or MDV-infected cells; 5 and 6, lysates from 358- labeled, PAA-teated, mock- or MDV-infected cells; 7 and 8, lysates from 353- labeled, unteated, mock- or MDV-infected cells. ICS, which contain antibodies to PAA sensitive A antigen (gp57-65), were used for lanes 1 to 4. Anti-pp38 was used for lanes 5 to 8. (C) pp38 is a serotype 1 specific antigen. Lanes: 1, lysate fi'om 35S-labeled MDV-infected cells; 2, lysate from 38S-labeled mock- infected cells; 3, lysate from 35S-labeled HVT-infected cells. (D) ' pp38 is expressed in latently infected lymphoblastoid tumor cells. Lanes: CON, lysate from 35S-labeled mock-infected cells; NF, lysate from 35S-labeled MDV- infected cells; MSB, lysate fi'om 35S-labeled MSB-l lymphoblastoid tumor cell line. (B) ICS contain antibodies reactive with pp38. Lanes: 1 and 3, lysates from mock-infected cells; 2 and 4, lysates from MDV-infected cells. The antisera used were anti-pp38 for lanes 1 and 2 and ICS for lanes 3 and 4. mm EM3 1 351 in st D1 Cd in 1% Fig. 8. A. pfi38 Is B. Effect of PAA on pp38 P osphorylated. Synthesis. 1 2 3 4 1 2 3 4 5 6 7 8 M INF - + — + INF - + - + — + — + (kd) -+~ PAA +-+ .--. +-+ - - -91 e4 -68 pp38- " . -- A Ag- - -43 a m... . e -29 p24— 1: p24- ~~ -18.4 C. pp38 Is D. pp38 Is Expressed E. pp38 Is an Serot¥pe—1 in MSB-l Cell in Immunogen in Speci 1c. the Absence of Birds with IUdR Treatment. MD 1 2 3 C I M 1 2 3 4 HVT - — + o it s - -+ — -+ INF MDV + - - N F B pp38- . . ” -pp38 . ' —pp38 -p24 —p24 p24- 92 and 4), the corresponding culture media from teated and unteated DEF cells were irnmunoprecipitated with immune chicken sera (ICS) to determine the amount of gp57-65 (A antigen), an MDV homolog of herpes simplex virus gC, which was known to be predominantly present in the media (10) and which is encoded by a gene that is sensitive to PAA teatment (14, 33a). The results of SDS-PAGE analysis showed that the gene encoding pp38 is relatively insensitive to PAA teatment (Fig. 8B, compare lane 6 with lane 8), while the gene encoding the A antigen is sensitive to PAA teatnent (Fig. 8B, compare lane 2 with lane 4). A preliminary Southern blot analysis suggested that no gene homologous to the MDV pp38 ORF is present in HVT (data not shown). However, since the HVT counterpart of an MDV antigen often shares a common epitope(s) (10, 16), irnmunoprecipitation analysis of mock- and HVT-infected DEF cell lysates, along with an MDV-infected cell lysate as a positive contol, was performed with anti- pp38 to further determine whether there is an HVT homolog. No HVT homolog of the MDV pp38 was found (Fig. 8C, lane 3). To directly demonstate whether a gene related to the pp38 gene identified in this study is expressed in the latently infected and tansformed MSB-l lymphoblastoid tumor cell line, we performed irnmunoprecipitation analysis of a lysate from 355-1abeled MSB-l cells (Fig. 8D, lane MSB), along with lysates from 36S-labeled mock— and MDV-infected DEF cells (Fig. 8D, lanes CON and NF) with anti-pp38. pp38 was expressed in the MSB-l tumor cell line in the absence of induction by IUdR (Fig. 8D, lane MSB). To determine whether pp38 is an irnmunogen in birds with MD, we used ICS for irnmunoprecipitation and SDS-PAGE analysis of lysates from mock- and MDV-infected DEF cells. ICS were found to contain antibodies reactive with pp38 (Fig. 8B, lane 4). This result was also confirmed by Western blot analysis, 93 in which ICS detected both the tpE-pp38 fusion protein and pp38 in an MDV- infected DEF cell lysate (data not shown). 94 DISCUSSION In this study, the use of cDNA analysis to detect the entire gene encoding pp38 was done as part of laboratory's general MDV gene identification program and, more specifically, resulted from an attempt to further locate the gene encoding the MDV B antigen (gp100, gp60, and gp49), which was originally thought to be located in the same region of the genome (32). This identification of the complete gene encoding pp38 and the subsequent analysis of it nucleotide sequence facilitated many experiment and observations that confirmed and greatly extended our understanding of this important MDV antigen. Of cental importance was the ability to prepare a highly specific antibody reactive against pp38 (anti-pp38) by using the gene nucleotide sequence and predicted amino acid sequence data in a fusion protein approach to antibody production. On the basis of analysis of the pp38 gene nucleotide sequence data and the single protein tuly irnmunoprecipitable by anti-pp38, the following point can be sununarized. Despite the early report of multiple phosphoproteins irnmunoprecipitable with monoclonal (6, 13, 22, 31) and polyclonal (6) antibodies, it appears from this study that pp38 is the primary gene product. Furthermore, it appears that p24 is not a processing or degradation product of pp38. Through the use of anti-pp38, it was possible to determine that the product of the gene described in this study is phosphorylated, serotype 1 specific (with respect to serotype 3), and expressed both in lytically infected fibroblast cells and in latently infected and tansformed lymphoblastoid tumor cells, thereby confirming the reported properties of pp38 tan a gene-based perspective. Also this is the first report that pp38 is an irnmunogen in birds with MD. Finally, by finding the complete gene encoding pp38, it was possible to determine that the gene encoding the B antigen complex is not located in the same region of the genome, as was previously 95 thought to be the case. When appropriate, the point summarized above will be discussed below to further contribute to our understanding of the molecular biology of MDV. To ensure that the pp38 identified in this study is the same as the 38 kDa phosphoprotein reported by others (6, 13, 31), we characterize the properties of pp38. Most of the properties of pp38 described here, such as its phosphorylation and failure to be detected in HVT-infected cells, are identical to those of the 38-kDa phosphoprotein, suggesting that these proteins are the same. The approach used to identify the gene encoding pp38 in this study was based on finding the ORF first and then determining it product, an approach entirely difierent from that of Cui et al. (6). In addition, nucleotide sequence analysis of the complete gene found by their approach (18a) resulted in an ORF nucleotide sequence identical to that of the pp38 ORF reported here. The term pp38 used in this study is derived from that which was first used by Cui et al. (6) and which was apparently based on the original discovery of a 38 kDa polypeptide as one of a group of three (31) and is consistent with the proposed MDV protein nomenclature (30). Furthermore, the size estimate of 38 kDa in this study (Fig. 6) is in full agreement with the published size (6, 31). Nakajirna et al. (22) used the term pp39/36, possibly as an extension of their earlier report of four polypeptides (13). However, the band more recently identified as pp39/36 (22) seems to be a single polypeptide, not a doublet, and most likely is the pp38 described by Cui et al. (6) and in present report. Of importance are the observations that pp38 (pp39/36) is always the most prominent (8 to 10-fold more) polypeptide of the group of phosphoproteins (6, 13, 22, 31) and that a larger polypeptide (41 or 43 kd) was not detected in this study. The protein data of this study (discussed in more depth below) all point to pp38 as the sole phosphoprotein among the groups reported by others. 96 Identification and sequence analysis of the gene encoding pp38 in this study provide further support for continuing the present nomenclature and are expected to provide clarity as MDV protein nomenclature evolves to accommodate gene identification (30). The fact that pp38 is expressed in the MSB-l lymphoblastoid tumor cell line (Fig. 8D) and tumor lesions of MDV-infected chickens (21) and that convalescent-phase sera from chickens with MD are capable of imrnunoprecipitating pp38 from MDV-infected DEF cells (Fig. 9, lane 4) fulfills the early criteria for a tumor antigen as originally set forth for polyomavirus large T antigen (11). Therefore, on the basis of these criteria, pp38 can appropriately also be referred to as an MDV tumor antigen. Later, the polyomavirus large T antigen was found to be an oncogene (9). Nakajima et a1. (22) suggested that the group of phosphoproteins may be associated with MDV oncogenicity. However, on the basis of the protein sequence homology analysis in this study, there is no significant homology between the predicted amino acid sequences of the pp38 gene and that of any known oncogenes. In addition, pp38 was not found to be a nuclear DNA binding protein (18a), although that is not a universal requirement for an oncogene. Silva and Lee (31) reported that p38 was detected in cells infected with the attenuated, avinllent serotype 1 Md11/76C MDV stain. Thus, if pp38 plays a role in MDV tansformation two possibilities exist: (i) it was modified during attenuation or (ii) it is necessary but insuficient and indirect, and attenuation involves another gene (3). Whether pp38 is a unique MDV oncogene remains to be determined. On the basis of their initial analysis of Agtll clones, Cui et al. (6) suggested that duplicate copies of the pp38 gene may exist in BamHI-D and -H, which contain MDV long terminal repeat (TRL) and long internal repeat (IRL) sequences, respectively. Since the 5’ end of the pp38 ORF identified in this 97 study is located in IRL, duplication of this part of the ORF in TRL region was considered a possibility. A 2.7-kbp DNA fragment which spans the long unique (U L) and TRL in BamHI-D was sequenced (data not shown). Duplication of the complete pp38 ORF in BamHI-D was not observed, but part of it (encoding 45 amino acids) was found because BamHI-D contains TRL. Previously, Sithole et al. (32) suggested that the gene for the MDV B antigen complex (gp100, gp60, and gp49) is located in BamHI-H, spanning the IRL-UL junction. Finding the pp38 ORF spanning the junction precludes finding the other gene in the same BamI-II-H region. The gene encoding the MDV B antigen complex has recently been identified, and it is not located in BamHI-H. The MDV homolog of the herpes simplex virus gB gene, previously reported to be located in BamHI-I3 and -K3 by Ross et al. (26), encodes glycoprotein gp100, gp60, and gp49 (5a), previously identified as MDV B antigen complex (16). In Northern blot analysis, a 1.95-kb tanscript for the pp38 gene was identified with both cDNA 2 and the EcoRI c subfragment of BamHI-H as probes (Fig. 2A and 2B). Considering that a 100-200 base poly(A) tail is usually present in mRNA (20), the size (1.95-kb) of the tanscript identified is consistent with the size of 1.85 kb as determined by S1 nuclease analysis (Fig. 3). Two minor tanscript of 1.3 and 0.8 kb were detected by the cDNA 2 probe; they were not detected by the EcoRI c subfragment probe. These result indicate that the two minor mRNAs are tanscribed from the EcoRI b subfragment, an observation that is similar to previous ones (3). Since the coding region of pp38 was located almost entirely (all but 3 nucleotides) in the EcoRI c subfragment, no conunon amino acid sequence is expected between the pp38 gene and the genes that are tanscribed to yield the 1.3 and 0.8 kb tanscript, should they encode any protein. 98 The calculated molecular weight of the polypeptide predicted for the pp38 gene is 31,169, a molecular weight which appears smaller than the apparent molecular weight of pp38 in SDS-PAGE. Three factors may contibute to the discrepancy. Extensive phosphorylation of pp38 (Fig. 8A) changes the secondary stucture and surface charges and subsequently slows the mobility of the protein (7). Stong hydrophilic domains predicted by hydrophilicity analysis (data not shown) and further suggested by antibody production against the trpE fusion proteins may affect the mobility of pp38 (18). N-linked glycosylation is an unlikely explanation for this size discrepancy, since there was no size change as a result of tunicamycin teatment (data not shown). Nakajirna et al. (22) reported that p24 is an MDV gene product and suggested that p24 either shares an overlapping DNA sequence with the pp38 gene (their p39/36) or is tanslated from a separate region with partial DNA homology to that of pp38 (their p39/36). In SDS-PAGE analysis, both pp38 and p24 were found in irnmunoprecipitates formed with anti-pp38 in this study, with the former being much more abundant. However, in this study pulse-chase analysis suggested and Western blotting confirmed that only pp38 is the gene product of the pp38 ORF. This conclusion is further supported by the observation that, among the multiple phosphoproteins, only pp38 was irnmunoprecipitated from a cell-free tanslation product by ICS (32a). These data indicate that no common epitope exist between pp38 and p24, despite the fact that these proteins are present in the same irnmunoprecipitates formed with anti-pp38. The nucleotide sequence analysis done in this study showed that there is no ORF overlapping the pp38 ORF which could potentially encode p24. Although one-third of the nucleotide sequence of the 1.95-kb tanscript of the pp38 gene is located in the IRL region of BamHI-H, only 15% of the nucleotide sequence of the ORF of the pp38 gene (the amino acid coding region) is in this 99 IRL region. While the same sequence was found in the TRL region of BamHI-D, as expected, the small size of the sequence (46 amino acids) precludes it encoding p24. Several questions need to be determined for future experiment: (1) whether or not p24 forms a specific complex with pp38 or is nonspecifically tapped during irnmunoprecipitation and (ii) whether p24 is MDV specific or cell specific. Cui et al. (6) reported that pp38 and four other proteins (135, 41, 24, and 20 kDa) were irnmunoprecipitated by antisera against two MDV-lacZ fusion proteins. Since the pp38 gene reported here is located in the same region (Fig. 1) as their two Jigtll clones (6), it is possible that the four other proteins are not encoded in that region of BamI-II-H. Since the lacZ system was used in their study (6), as opposed to the trpE system that was used in this study, it is possible that antibodies against the epitopes of the larger lacZ protein cross- react with some viral proteins. However, it is more likely that the four other polypeptides were nonspecifically tapped during irnmunoprecipitation. The fact that pp38 gene expression is relatively insensitive to PAA teatment suggest that pp38 could be one of the early classes of herpesvirus proteins. Furthermore, such an interpretation of the data is consistent with a previous report (29) which showed that an unidentified 1.9-kb E gene tanscript was detected in the BamI-II-H. While the precise location of this unidentified IE gene remains to be clarified in relation to the location of the pp38 gene, the distinct possibility exist that these genes are the same. Many IE genes in other herpesvinises tend to regulate the early or late gene expression (9). However, since the amount of pp38 was partially reduced when PAA was used, it is also possible that the gene encoding pp38 is a leaky late gene. Whether the pp38 gene is actually an E, early, or leaky late gene and whether it has a regulatory function remain to be determined. 100 ICS were found to contain antibodies capable of irnmunoprecipitating pp38 (Fig. 8B, lane 4), indicating that pp38 is an irnrnunogen in birds with MD. This result is difi‘erent from that previously reported (31). ICS also contain antibodies against the B antigen complex (16), which were found to elicit virus neutalization (15) and immunoprotection against MDV (24). Whether pp38, as an irnmunogen in birds with MD, is capable of eliciting virus neutalization and immunoprotection against MDV remains to be determined. The finding in this study that pp38 is expressed in the MSB-l cell line confirms previous observations (6, 13). However, IUdR, a drug known to enhance overall gene expression in the latently infected MSB-l lymphoblastoid tumor cell line (8), was used in the previous studies (6, 13) to detect pp38, implying that pp38 is not readily expressed in the latency stage. In this study, no IUdR was used and an abundant level of pp38 was expressed and detected in the latently infected MSB-l lymphoblastoid tumor cell line (Fig. 8D). The fact that pp38 is abundantly expressed in a constitutive manner not only in lytically infected fibroblast cells, but also in latently infected MSB-l lymphoblastoid tumor cells suggest that it may play a role in the establishment and maintenance of latency. 101 AGEIOWLEDGMI-INTS This research was supported primarily by a contact with Intervet International. We thank Quentin McCallum and Ruth A. Stinger of Michigan State University and I. A. I. Claessens of Intervet International for excellent technical assistance and Iohn C. Baker, Peter Brunovskis, Richard C. Schwartz, and Robert F. Silva for their careful reviews of the manuscript and many helpful suggestions. The pATH vectors used in this study were kindly provided by Richard C. Schwartz. We thank Lucy F. Lee (U SDA-ADOL) for sharing data prior to publication. 102 REFERENCES Aklyama. Y.. and S. Kato. 1974. Two cell lines from lymphomas of Marek’s disease. Biken I. 17:105-116. Bonner. W. M., and R. A. Luby. 1974. A film detection method for titiurn labeled proteins and nucleic acids in polyacrylamide gels. Eur. I. Biochem. 46:83-88. Bradley,G.,M.Hayashi.G.Lancx.ATanaka.andMNonoyama.1989. Stucture of the Marek’s disease vinis BamHI-H gene family: genes of putative importance for tumor induction. I. Virol. 63:2534—2542. Calnek. B. K. 1985. Marek’s disease - a model for herpesvinis oncology. Crit. Rev. Microbiol. 12:293-320. Chen. 1., and L. F. Velicer. 1991. Multiple bidirectional initiations and terminations of tanscription in the Marek‘s disease vinls long repeat regions. I. Virol. 65:2445-2451. Chen. 1., and L. F. Velicer. 1991. The Marek's disease vims B antigen complex (gplOO, gp60, gp49) is the homolog of the herpes simplex virus glycoprotein B (gB). Summitted for publication. Gui, 2., D. Yen. and L. F. Lee. 1990. Marek’s disease virus gene clones encoding virus-specific phosphorylated polypeptides and serological characterization of fusion proteins. Vinls Genes 3:309-322. DeCaprlo. I. A., I. W. Ludlow, D. Lynch, Y. Fin'ukavva. I. Grifin. H. Pivvnica-Worms, C.-M. Huang, and D. M. Livingston. 1989. The product of the retinoblastoma susceptibility gene has properties of a cell cycle regulatory element. Cell 58:1085-1095. Dunn, E, and K. Naserian. 1977. Induction of Marek’s disease virus antigens by IdUrd in a chicken lymphoblastoid cell line. I. gen. Virol. 34:413-419. 10. 11. 12. 13. 14. 15. 18. 17. 103 F1elds,B.N..D.M.Knipe.R.M.Chanock.MS.Hersch.I.L.Melniclt.T. P. Monath. and B. Roizman. 1990. Herpes simplex virus and it replication, p. 1795-1841. B. Roizman, and A. Sears. Fields Virology. 2nd edition. Raven Press, New York Mg". C., K. Naserlan. and L. F. Velicer. 1983. Marek’s disease herpesvirus. IV. Molecular characterization of Marek’s disease herpesvirus A antigen. I. Virol. 45:1228-1234. Habel. K 1965. Specific complement-fixing antigens in polyoma tumors and tansformed cells. Virology 25:55-61. Hopp, T. P., and K R. Woods. 1981. Prediction of protein antigenic determinant from amino acid sequence. Proc. Natl. Acad. Sci. USA 78:3824-3828. llmta.K,KNahIima.MNaito,S.ILArm.S.Ueda.S.Kato.andKHirai. 1985. Identification of Marek’s disease vims-specific antigens in Marek’s disease lymphoblastoid cell lines using monoclonal antibody against virus-specific phosphorylated polypeptides. Int. I. Cancer 35:257-264. Ilulta. K., S. Ueda. S. Kato, and K Hirai. 1983. Monoclonal antibodies reactive with the surface and secreted glycoproteins of Marek’s disease virus and herpesvirus of turkeys. I. gen. Virol. 64:2597-2610. Ikuta. K., S. Ueda. S. Kato, and K Hirai. 1984. Processing of glycoprotein gB related to neutalization of Marek’s disease virus and herpesvirus of turkeys. Microbiol. Immunol. 28:923-933. Isfort, R. I., I. Sithole, H.-I. Eng, and L. F. Velicer. 1986. Molecular characterization of Marek’s disease herpesvirus B antigen. J. Virol. 59:411- 419. Konopka.I.B.,R.L.Davis,S.MWatanabe,A.S.Ponticelli.L.Schifi- Malnr, N. Rosenberg, and O. N. Witte. 1984. Only site-directed antibodies 18. 18a. 19. 21. 104 reactive with the highly conserved arc-homologous region of the v-abl protein neutalize kinase activity. J. Virol. 51:223-232. LahlIani.R.S..E.W.Otteson.I.D.Adlish.andS.C.S.Ieor. 1991. Characterization of a human Cytomegalovirus 1.6-kilobase late mRNA and identification of it putative protein product. J. Virol. 65:373-381. Lee, L. F. personal communication. Many, T., M. Malldson. and Y. Becker. 1988. RNA tanscript of Marek’s disease vims (MDV) serotype-l in infected and tansformed cells. Virus Genes 2:49-68. Monte11,C.,E.F.Fisher,MH.Canithers, andA.I.Berk.1983. Inhibition of RNA cleavage but not polyadenylation by a point mutation in mRNA 3’ consensus sequence AAUAAA. Science 305:600-605. Naito,M.,KNakaiima.N.Ivva.KOno,I.Yoshida.T.Konobe,KIkuta,S. Ueda. S. Kate. and K Hirai. 1986. Demonstation of a Marek's disease virus-specific antigen in tumor lesions of chickens with Marek’s disease using monoclonal antibody against a vims phosphorylated protein. Avian Pathol. 15:503-510. NWLKMMNaito,S.Ueda.S.Kato,andKHIrai. 1987. Analysis of Marek’s disease virus serotype l-specific phosphorylated polypeptides in virus-infected cells and Marek’s disease lymphoblastoid cells. I. gen. V'irol. 68:1379-1389. Naserlan. K, and L. F. Lee. 1976. Selective inhibition by phosphonoacetic acid of MDV DNA replication in a lymphoblastoid line. Virology 74:188- 193. Ono.K,M.Ta1tssliima.T.Ishikavva.M.Hayashi.I.Yoehida. T.Konobe,K Ikuta. K Nakajirna. S. Ueda. S. Kato. and K Hirai. 1984. Partial protection against Marek's disease in chickens immunized with glycoproteins gB 31. 105 purified from turkey-herpesvinis-infected cells by afinity chromatography coupled with monoclonal antibodies. Avian Dis. 29:533-539. Payne. L. N. 1985. Marek's disease - scientific basis and methods of contol. Martinus Nijhoff Publishing, Boston. Rms,L.I.N..M.Sanderson.S.D.Soott.M.M.Birms,T.Doel.andB. Wm. 1989. Nucleotide sequence and characterization of the Marek’s disease virus homologue of glycoprotein B of herpes simplex vims. I. gen. Virol. 70:1789-1804. Sambrook. I., E. F. Fritch. and T. Maniath. 1989. Molecular cloning: a laboratory manual, 2nd edition. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. Sanger, F., S. Nlcltlen. and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. Schat. K A., A. Bucltmaster. and L. I. N. Ross. 1989. Partial tanscription map of Marek’s disease herpesvinis in lytically infected cells and lymphoblastoid cell lines. Int. I. Cancer 44:101-109. Schat.KA.,G.F.deBoer,KHirai.L.F.Lee,KNakinma.R.F.Silva.and L. F. Velicer. 1988. A proposed nomenclature for the protein of Marek’s disease herpesvirus, p. 438-443. In S. Kato, T. Horiuchi, T. Mikami, and K. Hirai. (ed.), Advances in the Marek’s disease research. Iapanese association on Marek’s disease, Osaka, Iapan. Silva. R. P., and L. F. Lee. 1984. Monoclonal antibody-mediated irnmunoprecipitation of proteins from cells infected with Marek’s disease vims or turkey herpesvinis. Virology 136:307-320. Sithole. I., P. M. Coussens, and L. F. Velicer. 1988. Identification of Marek’s disease herpesvirus B antigen’s precursor polypeptide and the gene encoding it, p. 148-154. In S. Kato, T. Horiuchi, T. Mikami, and K. 106 Hirai. (ed.), Advances in the Marek's disease research. Iapanese association on Marek's disease, Osaka, Iapan. 32a Stringer, R. A., K Chen. and L. F. Velicer. Unpublished data. 33. Sugaya. K. G. Bradley. M. Nonoyama. and A. Tanaka. 1990. Latent tanscript of Marek's disease virus are clustered in the short and long repeat regions. I. Virol. 64:5773-5782. 33a. Velicer, L. F. Unpublished data. 34. Weber, K, and M. Osborn. 1969. The reliability of molecular weight determination by dodecyl-sulfate-polyacrylamide gel electophoresis. J. Biol. Chem. 244:4406-4412. ChapterIV 'IheMarek'sDiseueVirusBAntigenGomplex(gp100,gp60.gp49) IstheHomologoftheHerpesSimplexVirmGyooproteianB) Xinbin Chen, and Leland F. Velicer. 1991. Submitted for publication. 107 108 ABSTRACT Marek’s disease (MD) is caused by Marek’s disease vims (MDV), and is prevented by immunization with an antigenically related, but apathogenic, herpesvirus of turkey (HVT). Among the major glycoproteins found in MDV- and HVT-infected cells is the B antigen complex (gplOO, gp60 and gp49), detected by irnmunoprecipitation and SDS-PAGE analysis with antisera previously shown to be reactive with B antigen in immunodifiusion analysis. However, the gene encoding this important MDV antigen complex was not unequivocally identified. Recently, an MDV homolog of the gene encoding herpes simplex vims (HSV) glycoprotein B (gB) was identified and sequenced (Ross, L. I. N., M. Sanderson, S. D. Scott, M. M. Binns, T. Doel, and B. Milne. I. gen. Virol. 70:1789-1804. 1989). To determine if the MDV gB homolog gene might encode the B antigen, antisera were prepared against tpE fusion proteins of the MDV gB homolog (tpE-MDV-gB). These antisera irnmunoprecipitated gp100, gp60, gp49 and pr88. Based on size comparison, tpE-MDV-gB competition and blocking assays, it was concluded that: l) the MDV homolog of HSV 98 gene encodes MDV B antigen, and 2) pr88 was found to be the only precursor polypeptide, while pr44, previously thought to dimerize to form pr88, is a protein that is nonspecifically tapped during irnmunoprecipitation. The antisera against tpE-MDV-gB also contained antibody reactive with the HVT B antigen, consistent with the known antigenic relatedness between the MDV and HVT B antigens. A new MDV glycoprotein, gp98, was also detected by the antisera. Result from pulse-chase analysis suggest that gp98 is a glycoprotein processing intermediate. The MDV gB homolog processing pathway appears to involve cotanslational glycosylation of pr88 to form gp98, which is further processed to yield gp100, which is then cleaved to form gp60 and gp49. This 109 processing pathway is consistent with those of other gB homologs, further supporting the gene identification described above. 110 INTRODUCTION Marek’s disease (MD) is a lymphoproliferative disease of chickens caused by cell-associated Marek's disease herpesvinis (MDV), and is characterized principally by T-cell lymphomas and peripheral nerve demyelination (4, 24). The disease has been efl'ectively contolled by vaccination with apathogenic but antigenically related herpesvirus of turkey (HVT). To determine the mechanism of immunity conferred by HVT, efiort have focused on antigens found in common between MDV and HVT, particularly the A and B antigens (9, 10, ll, l5, l6, 17, 18, 29, 32). The genes encoding the common MDV and HVT A antigens (gp57-65) have been identified and characterized; these represent the MDV/HVT homologs of herpes simplex vims (HSV) glycoprotein C (gC) (9, 10). The MDV and HVT B antigens were characterized as a complex of three glycoproteins, gp100, gp60 and gp49 (l6, 17, 29, 32). Tunicamycin (TM) inhibition of N-linked glycosylation resulted in identification of various unglycosylated, or possibly O-linlted glycosylated, precursor polypeptides: pr88/pr83 of MDV and pr90 of HVT (l6), pr80/pr110/pr125 of MDV (11), or pr88 and pr44 of MDV and HVT (17, 32). In the latter report, it was hypothesized that pr44 dimerizes to form pr88 (32). HVT B antigen complex prepared by immune afinity purification was also shown to elicit partial protective immunity against MDV in chickens (23), and antibodies against B antigen are able to neutalize the infectivity of MDV in cell cultures (16). A preliminary report by Sithole et al. (31) suggested that the gene encoding MDV B antigen was located in the BamHI-H fragment, which spans the junction of the unique long (UL) and inverted repeat long (IRL) regions of the MDV genome. Upon further analysis, Chen et al. (5) reported that a 111 phosphoprotein pp38 was encoded in this region of BamHI-H; moreover, no open reading frames characteristic of glycoprotein genes were identified. Thus, the location of the gene encoding MDV B antigen remained to be determined. Ross et al. (26) reported the identification of the gene encoding the MDV homolog of HSV glycoprotein (gB) (MDV gB homolog gene). By Western blot analysis, the MDV gB homolog was found to be composed of gpllO, gp64 and gp48 (26). In the presence of TM, three smaller polypeptides (94, 90 and 84 kDa) were found to represent the unglycosylated, or possibly O-linked glycosylated, precursor polypeptides, with a 48 kDa polypeptide being a tuncated form of the former ones (26). Solely based on the estimated sizes of the three glycoproteins, they suggested that MDV gB homolog could be the MDV B antigen complex. However, despite the fact that antisera against the B antigen complex neutalize MDV infectivity (16), their anti-peptide sera against MDV gB homolog failed to neutalize MDV infectivity (26), an observation which is not consistent with their above suggestion. It was suggested (26) that this failure to neutalize MDV could be explained if their anti-peptide sera reacted predominantly with the primary stucture of the antigen, and therefore may be unable to react with epitopes on the intact virion. This study provides conclusive evidence that the MDV B antigen complex is in fact the MDV homolog of HSV gB. In addition, it was also learned that pr88 is the actual unglycosylated (or possibly O-linked glycosylated) MDV gB primary precursor polypeptide; and that MDV gB homolog processing has a similar pattern as to those of equine herpesvinls (EHV), cytomegalovirus (CMV), pseudorabies vinls (PRV) and varicella-zoster vinls (VZV). 112 MATERIALS AND METHODS Cells and viruses. The preparation, propagation and infection of duck embryo fibroblast (DEF) cells with MDV or HVT was performed as described previously (6, 13). The MDV GA stain used in this study was at cell culture passage level 6 following isolation of cell-free vims from feather tips obtained from infected birds with MD. The HVT FC-126 stain of vaccine virus at cell culture passage level 13 was used in this study. Antisera. The preparation and characterization of rabbit antibody against MDV-infected DEF cell plasma membrane (RorPM) has been described previously (17, 32). RaPM and the monoclonal antibody IAN86 are irnmunoreactive with the MDV B antigen complex (17, 29, 32). GenerationoftpEfuslonproteiruofMDVgBhomolog (tpE-MDV-gB). The vector system used to express the MDV gB homolog gene’s open reading frame (ORF) in Escherichia coli consist of a group of plasmids (pATH vectors) encoding approximately 37 kDa of the bacterial tpE ORF under contol of the inducible tp operon promoter (20). A polylinker with multiple cloning sites at the 3’ end of the tpE ORF allows in-frame insertion of foreign ORFs. Three DNA fragment, designated as fragment 1, 2 and 3 (Fig. 1), from the MDV gB homolog gene’s coding region (26), were purified and cloned into various tpE expression vectors according to the method by Sambrook et al. (27), and their respectve tpE-MDV-gB fusion proteins were generated. Fusion protein antibody production. Fusion proteins were purified by electophoresis in preparative 7.5% SDS-polyacrylamide gels. The gels were stained with Coomassie brilliant blue to identify the positions of the fusion proteins. The fusion proteins were excised and emulsified in complete Freund’s adjuvant (Difco Laboratories). Initial immunization of 2.0 kg New Zealand White female rabbit was with 200 to 500 pg of proteins. Rabbit were boosted after 3 113 FIG. 1. The locations and sizes of DNA tagment used to generate tpE fusion proteins. Fragment l, a 566-bp XbaI-EcoRV subfragment of BamHI-I3, whose tpE fusion protein was used to generate antiserum Z43. Fragment 2, an 1,058-bp EcoRV-BamHI subfragment of BamHI-I3, whose tpE fusion protein was used to generate antisenim Z45. Fragment 3, a 277-bp BamI-II-HaeIII subfragment of BamI-II-Ka, whose tpE fusion protein was used to generate antisenim 242. The locations of the signal peptide, putative protease cleavage site, and tansmembrane and cytoplasmic domains were reported by Ross et al. (26). 114 Fig. 1. Fragment 1. Amino Acid 59 to 247 Fragment 2. Amino Acid 248 to 601 Fragment 3. Amino Acid 602 to 693 I t l 1 l 1 1 A a I ran zee see «so see see 700 see I l | I I signal Peptide Putative Protease Transmembrane and from Residues Cleavage Site Cytoplasmic Domains 1 to 21 between Residues from residues 434 to 435 683 to 865 115 weeks with 100 to 200 pg of proteins in incomplete Freund’s adjuvant, and sera were collected 7 to 10 days later. Subsequent booster immunizations were approximately 2 weeks apart, as needed to obtain detectable and usable titers. The antisera against the tpE fusion proteins of fragment 1, 2 and 3 (Fig. 1) were designated as 243, 245, and 242, respectively. Radiolabelingdproteimandpreparationoflysates. Mock-,HVT- and MDV-infected DEF cells were labeled with [358] methionine (specific activity, 1,000 pCi/mmol, ICN) at 48 h postinfection for 4 h, as previously described (13). Pulse-chase labeling was performed as previously described (18, 32), with the following modifications. Cells were preincubated without methionine for l h; pulse labeling was done by incubation with 250 p01 of [3551 methionine per 60 mm plate in 1.0 ml media for 5 min; and chases were done in the presence of unlabeled methionine for 0, 5, 10, 20, 40, 80, 160, and 320 min. For labeling in presence of 2.0 pg/ml tunicamycin (TM) (Calbiochem-Behring Corp.), cells were preincubated with TM for l h, and teated at same concentation during the 4 h labeling period. After labeling was complete, cells were washed three times with ice-cold phosphate-buffered saline, followed by the addition of ice-cold detergent buffer (0.01 M NaHzPO4-Na21-IPO4, pH 7.5, 0.1 M NaCl, 1% Triton X- 100, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate) (36) to lyse cells. The lysates were clarified by centifugation as previously reported (36). Imrtunopreclpltatlon and SDS-PAGE analysis. Irnmunoprecipitation analysis was carried out as previously described (13). The irnmunoprecipitates I were washed, suspended in sample buffer, and analyzed by sodium dodecyl sulfate-polyacrylamide (7 .696) gel electophoresis (SDS-PAGE). Protein markers (BRL) were used as molecular mass standards. Molecular sizes were calculated by interpolation between standard proteins by the method of Weber and Osborn (34). Fluorography was carried out as previously described (1). 116 RESULTS IdentificationofMDVandHVTgBhomologgeneproduct. The antisera against tpE-MDV-gB (243, 245 and 242), along with the monoclonal antibody, IAN86, and polyclonal antiserum, RaPM, were used for irnmunoprecipitation analysis of lysates from both MDV- and HVT-infected DEF cells labeled with [358] methionine in the absence or presence of TM. In the absence of TM, gp100, gp60 and gp49 were detected by IAN86 and RaPM in both MDV- and HVT-infected cell lysates, as expected (Fig. 2, lanes 1, 3, 5, and 7). Note that for HVT, the gp60 and gp49 counterpart migrate slightly difi'erent from MDV’s (Fig. 2, compare lane 1 with lane 3). However, established MDV terminology (29) is used here for simplicity and ease of identification. Also note that the IgG heavy chain of polyclonal RaPM tends to distort the migration of gp49 (Fig. 2, compare lanes 5 and 7 with lanes 1 and 3), a result of the volume of this lower titer serum needed for irnmunoprecipitation. For MDV-infected DEF cells, while little, if any, glycosylated MDV-specific proteins were detected by 242 (Fig. 2, lane 17); gp100, and an additional new MDV glycoprotein, designated gp98 (immediately below gp100), were detected by 243 and Z45 (Fig. 2, lanes 9 and 13). However little, if any, gp60 and gp49 were detected by 243 and 245 (Fig. 2, lanes 9 and 13). In contast, in HVT-infected DEF cells, gp60 and gp49 were detected only by Z45 (Fig. 2, lane 15), while gp100 was detected by 243, 245 and 242 (Fig. 2, lanes 11, 15 and 19). In the presence of TM, virus-specific pr88 was detected by all five antisera used (Fig. 2, Lanes 2, 4, 6, 8, 10, 12, 14, 16, 18, and 20). Note that the apparent molecular weight for the precursor polypeptides of HVT and MDV are approximately equal (Fig. 2, lanes 2, 6, 10, 14, and 18 compared with lanes 117 FIG. 2. Identification of MDV and HVT gB homolog gene product. Lysates from MDV- or HVT-infected DEF cells labeled in the absence or presence of TM (see Materials and Methods) were irnmunoprecipitated by IAN86, anti-PM (RaPM), Z43, Z45, and Z42 as indicated above each lane. IAN86 and anti-PM (RaPM) were used as positive contols, which were known to be able to irnmunoprecipitate the MDV B antigen complex (gp100, gp60, gp49) (lanes 1, 3, 5 and 7) and it precursor polypeptide pr88 (lanes 2, 4, 6 and 8). gp98, a protein immediately below gp100 in lanes 9 and 13, is detected only in the irnmunoprecipitates of MDV-infected DEF cell lysates with antisera Z43 and Z45. Note that p79, a band below pr88, is an MDV-HVT-specific protein with a tendency to be nonspecifically tapped, and therefore is present in all lanes, except lanes 1 to 4 where the monoclonal antibody IAN86 was used. Virus- specific proteins are indicated to the left. Molecular mass standards are indicated to the right. Il+I--| I“) 4567891011121314151617181920 -++--++--++—-++-— +--++--++-—++--++ +-+-+—+-+—+-+—+-+ | anti—PM | 243 | 245 | 242 -. '9 Is - ’47 -3 -_='--.: -68 C. .- .-.-“ ..- 0- -43 -- 119 4, 8, 12, 16, and 20, respectively). Based on size comparison, it appears that the pr88 irnmunoprecipitated by the antisera against tpE-MDV-gB is the same pr88 irnmunoprecipitated by IAN86 and RaPM, suggesting that the MDV B antigen complex is the MDV gB homolog. The band that often appears below pr88 is p79, an MDV-HVT-specific protein with a tendency to be nonspecifically tapped (17). While it presence is nonspecific (in respect to the sera used) (17, unpublished result) and unintended, p79 often serves as a useful internal marker within the same gel lane; it presence below pr88 in all but the monoclonal antibody lanes (Fig. 2, lanes 1 to 4) serves to stengthen the size comparison refered to above. Previously, a polypeptide called pr44 was thought to be the precursor polypeptide for the B antigen complex (1 7, 32). However, in this experiment there is no virus-specific pr44 irnmunoprecipitated by antisera against tpE-MDV-gB (Fig. 2, lanes 10, 12, 14, 16, 18, and 20), as confirmed by subsequent fusion protein competition analysis. Although there is a band at approximately the 44 kDa position (the top band of the doublet) (Fig. 2, lanes 2, 4, 6, 8, 10, 12, 14, 16, 18, and 20; upon longer exposure two light bands were visible at lanes 2 and 4 where IAN86 was used), it now appears that it, probably what was previously termed pr44 (17, 32), is only one of the several nonspecific bands of that approximate size that are tapped during irnmunoprecipitation. That pr88 is the only precursor polypeptide for MDV gB homolog, is further confirmed by fusion protein competition analysis (see below). Consistent with the above suggestion that the MDV B antigen complex is the MDV gB homolog, antisera against tpE-MDV-gB should recognize ngO and gp49. To enhance the detection of these MDV gB homolog glycoproteins, boiling of MDV-infected cell lysates prior to immunoprecipitation was used to disrupt protein secondary stucture; with the intent that boiling might result in a 120 FIG. 3. Boiling of MDV-infected cell lysates to enhance the detection of MDV gB homolog by Z45. Boiled (lanes 4, 5, 6) or unboiled (lanes 1, 2, 3) MDV-infected cell lysates were irnmunoprecipitated by preirnmune serum (pie), Z45 and anti-PM (RaPM) as indicated above each lane. Virus-specific proteins are indicated to the left. Molecular mass standards are indicated to the right. 121 Lanes 1 2 3 4 5 6 Pre + — _ + _ .. Z45 - — + _ _ + Anti-PM — + — — + — Boiling - - - + + + I 913100" .8. a —97 p79- ‘ l s .00 gp49- ‘ -43 122 protein’s hidden continuous epitopes becoming available for recognition by antisera against tpE-MDV-gB. No MDV-specific proteins were detected by preirnmune sera, except for the nonspecifically trapped p79 (Fig. 3, lanes 1 and 3), whether the lysate was unboiled or boiled. In the Z45 irnmunoprecipitate of the boiled MDV-infected cell lysate (Fig. 3, lane 6), not only was more gp100 detected (Fig. 3, lane 6 compared to lane 3), but also gp60 and gp49 were now detected; the latter two were only minimally detectable in the irnmunoprecipitate of unboiled MDV-infected cell lysate (Fig. 3, lane 3 compared with lane 6), suggesting that the antisera against tpE-MDV-gB are mainly reactive with continuous epitopes. In contast, much less gp100, gp60 and gp49 were detected by RaPM when the lysate was boiled (Fig. 3, lane 2 compared with lane 5), suggesting that the majority of antibodies in the RaPM sera are reactive with conformational (discontinuous) epitopes that were disrupted by boiling. EtionprotelnoornpefitionanalysisofMDV-infectedcelllysates. To further assess the specificity of the antisera against tpE-MDV-gB, and to confirm that MDV B antigen is a MDV gB homolog, a competition assay was performed in which fragment 1, 2 and 3 fusion proteins were separately added to each of three boiled lysates prior to addition of their respective antisera, as indicated above each lane (Fig. 4A and 4B, lanes 5, 7 and 9) while all three fusion proteins were added to one boiled lysate when RaPM was used (Fig. 4A and 4B, lane 1). In the absence of TM, no virus-specific proteins were detected in MDV- infected cell lysate when preirnmune sera was used as a contol (Fig. 4A, lane 3). When RaPM and Z45 were used, trpE-MDV-gB competition resulted in gplOO, gp60 and gp49 no longer being detected (Fig. 4A, compare lanes 2 and 4 with lanes 1 and 5, respectively), further supporting the earlier suggestion (based on size comparison) that the MDV B antigen complex is a MDV gB 123 FIG. 4. Fusion protein competition analysis of MDV-infected cell lysates. (A) MDV-infected cell lysates were boiled prior to irnmunoprecipitation. Fragment l, 2 and 3 fusion proteins were added separately to each of three boiled lysates prior to addition of their respective antisera as indicated above each lane (lanes 5, 7 and 9). All three fusion proteins were added to one boiled lysate when anti-PM (RaPM) was used (lane 1). An identical amount of phosphate lysis buffer was added to the contols (lanes 2, 3, 4, 6 and 8). (B) All of the assay conditions were the same as in panel A except that the lysates were from MDV-infected DEF cells labeled in the presence of TM. The virus specific proteins are indicated to the left at each panel. Molecular mass standards are indicated to the right of panel B. P, representing preirnrnune sera. 124 A. Metabolic Labeling inAbsence ofTM Lanes 1 2 3 4 5 6 7 8 9 TrpE-gB + - - - + - + _, + Sera aPM | PI Z45 | 243 | 242 gplOO- p79- «- gp60- - ‘I- - gp49- . II 1 - 9- . B. MetabolicLabelinginPresenoeofTM 1 2 3 4 5 6 7 8 9 + - - - + — + - + aPiI | Pl 245 | 243 | 242 F”? ' t t -200 u-e -97 pr88: .. b” p79 '0'” .768 .... - ,u' I .. -43 125 homolog. When Z43 was used, gp100, gp60 and gp49 were detected in the boiled MDV-infected cell lysate (Fig. 4A. lane 6); the latter two were not detectable in the irnmunoprecipitate of unboiled MDV-infected cell lysate (Fig. 2, lane 9). TrpE-MDV-gB competition resulted in gplOO, gp60 and gp49 becoming undetectable (Fig. 4A. compare lane 6 with lane 7). While little, if any, gp100, gp60 and gp49 were detected by Z42 in unboiled lysate (Fig. 2, lane 17), all three were detectable when the lysate was boiled (Fig. 4A. lane 8). TrpE-MDV-gB competition resulted in gplOO, ngO and gp49 becoming undetectable (Fig. 4A. compare lane 8 with lane 9). Note that nonspecific tapping was increased, most notably of p79, when tpE-MDV-gB was used in the competition assay (Fig. 4A, compare lanes 4, 6, and 8 with lanes 5, 7, and 9, respectively). Result similar to those that occured with tpE-MDV-gB competition analysis were obtained by irnmunoprecipitation analysis of MDV- infected cell lysates by antisera (RaPM and Z45) that had been preincubated with tpE-MDV-gB (designated as tpE-MDV-gB blocking assay) (data not shown). In the presence of TM, pr88 was not detected by preirnmune sera (Fig. 4B, lane 3). While pr88 was detected in the irnmunoprecipitates of the boiled lysates when RaPM, Z45, Z43 and Z42 were used, it became undetectable as a result of tpE-MDV-gB competition (Fig. 4B, compare lanes 2, 4, 6, and 8 with lanes 1, 5, 7, and 9, respectively). A similar pattern (as in Fig. 4A) of increased nonspecific tapping was observed, especially of p79, when tpE-MDV-gB was used in the competition assay. Result similar to those that occured with tpE- MDV-gB competition assay were also obtained by a tpE-MDV-gB blocking assay similar to that described above for Fig. 4A (data not shown). Together, these result are consistent with the notion that pr88 is the primary precursor polypeptide of MDV gB homolog; this further support the above conclusion that the MDV B antigen complex is the homolog of HSV gB. EneticsofMDVgBhomologglycoproteinprooessingasdeterminedby pulse-chase analyst. For EHV (21, 33), CMV (2, 3), PRV (25, 35), and VZV (14, 22), a glycoprotein intermediate of gB homolog initially appears, which is slightly smaller in molecular mass than that of the mature glycoprotein. Upon further processing, the glycoprotein intermediate is converted to the mature glycoprotein, which is subsequently cleaved to form two smaller cleavage product. Since there is a 98 kDa glycoprotein detected by antisera Z43 and 245 (Fig. 2, lanes 9 and 13), as well as two smaller cleavage product gp60 and gp49 (16, 32), a similar processing pathway appeared to be involved in MDV gB homolog processing. To determine if the 98 kDa glycoprotein is an intermediate in the processing of pr88 to gp100, a 5 min pulse was done and was followed by various chase periods (Fig. 5). Since RcPM does not recognize gpQ8 as well as 245 does, but does recognize gp60 and gp49 when the lysate is not boiled (Z45 does not), a pool of these two antisera were used. Upon irnmunoprecipitation and SDS-PAGE analysis, the amount of gp98 gradually decreased, being easily detectable only with a 0 min chase (Fig. 5, lane 1) and barely detectable with a 5 min chase (Fig. 5, lane 2). Coincident with this decrease, the amount of gplOO gradually increased (Fig. 5, lanes 1 to 5). After a 20 min chase, cleavage of gplOO to form gp60 and gp49 began to take place, with most of this processing occuring between 40 and 80 min of chase (Fig. 5, lanes 5 and 6; upon longer exposure, gp60 and gp49 were visible in lane 5). As this processing continued gplOO was reduced to very low levels (Fig. 5, lanes 7 and 8, still seen in four-fold longer exposures of lane 8). Processing of gp60 and gp49 began to occur after a 80 min chase, with most of protein processing occuring between 160 and 320 min chases (Fig. 5, lanes 7 126 127 FIG. 8. Kinetics of MDV gB homolog processing as determined by pulse- chase analysis. A 5 min pulse labeling was followed by chases of various times shown above each lane. Irnmunoprecipitation analysis was done with a pool of antisera Z45 and anti-PM (RaPM). The virus specific proteins are indicated to the left. Molecular mass standards are indicated to the right. ififl D AA-U Lanes 12345678 Pulse(min) 5 Chase(min) 0 5 10 20 40 80 160 320 gp100= "' *---- -97 gp98’ 60 -68 99 ‘ .- gp49- 129 and 8). In summary, these data show that the 98 kDa glycoprotein is the intermediate between pr88 and gplOO, which is then cleaved to yield gp60 and gp49. 130 DISCUSSION This study provides several lines of evidence that the MDV B antigen complex (gp100, gp60, gp49) is encoded by the MDV gB homolog gene. This conclusion is based on irnmunoprecipitation analyses with antibodies to tpE- MDV-gB, and the use of tpE-MDV-gB competition and blocking assays. Further, size comparison of these three glycoproteins with those detected with anti-peptide sera prepared based on the sequence of the newly discovered MDV homolog of HSV gB. are consistent with this finding. Finally the glycoprotein processing pathway reported in this study is typical of those for gB homologs of other herpesviruses. This conclusion eliminates, not only the uncertainty that previously existed concerning the identity and location of the gene encoding the MDV B antigen complex, but also current confusion regarding the nomenclature of the MDV B antigen complex. The three glycoproteins (gplOO, gp60, gp49) first seen on SDS-PAGE (29) were previously shown to be the MDV B antigen complex (17) through use of antisera that was originally defined as having anti-B reactivity by immunodifiusion analysis, the historical basis for naming several antigens A., B and C (7, 8) found in common between MDV and HVT. Thus, the use of the letter B in MDV glycoprotein nomenclature predates the first use of gB . terminology in the HSV and other herpesvinis systems. Prior to the evidence reported in this paper, confusion sometimes arose when it was then erroneously assumed, based on the use of the letter B, that the MDV B antigen was a homolog of HSV gB. Compounding the MDV B antigen nomenclature confusion was the frequent use of the term gB by some MDV workers, as still another way to refer to the glycoprotein B antigen, not because of the evidence of homology to HSV gB. Their tentative nomenclature choice was to use gA and gB. for the secreted (A antigen) and virus neutalization-related (B antigen) glycoproteins 131 (15, 16). Now that homology of the MDV B antigen with HSV gB has been shown experimentally in the study reported here, it is suggested that the gene encoding it be designated as the MDV homolog of the HSV gB gene (MDV gB homolog gene), and that the glycoproteins (gp100, gp60, gp49) be designated as the MDV homolog of the HSV gB (MDV gB homolog). For current report, the tentative use of the abbreviation MDV gB seems appropriate, following the precedent of Ross et a1. (26); but that may be reconsidered as alphaherpesviruses and MDV/HVT glycoprotein nomenclature develops. Result from TM inhibition and irnmunoprecipitation analysis (Fig. 2, lanes 10, 14 and 18), tpE-MDV-gB competition assay (Fig. 4B) and tpE-MDV-gB blocking assay (data not shown) clearly demonstate that pr88 is the primary unglycosylated (or possibly O-Linked glycosylated) precursor polypeptide of the gp100, gp60 and gp49 complex. This interpretation is also supported by report from studies of gB homolog proteins of other herpesvinlses. In cases of EHV (21, 33), CMV (2, 3), VZV (14, 22), PRV (25, 35), and bovine herpesvirus (BI-IV) (12), where the mature gB homolog proteins are cleaved to form two smaller product, only one unglycosylated precursor polypeptide is observed; and no cleavage of the unglycosylated precursor polypeptide was reported. The appearence of a gp98 intermediate in this study (Fig. 2, lanes 9 and 13; Fig. 5, lane 1) is consistent with the appearence of a similar sized molecule at 0 and 3 min chase intervals after a 5 min pulse in an earlier study of B antigen processing (32). With identification of gpQ8 as an intermediate between pr88 and gplOO (Fig. 6), MDV gB homolog processing pathway appears similar to that observed for EHV (21, 33), CMV (2, 3), PRV (25, 35) and VZV (14, 22) gB homolog processing. The MDV gB homolog processing pathway is outlined in Fig. 6, and is based on a combination of the conclusions of this study and part of those 132 FIG. 8. MDV gB homolog processing pathway. 133 Fig. 6. p788 (unglycosylated precursor polypeptide) Block by tunicamycin 4 «e Cotranslational glycosylation gp98 (glycoprotein intermediate) <+ Further processing, possibly by addition of sialic acid gplOO (mature glycoprotein) Block by monensin a «e Protease cleavage gp60 m9 (cleavage products ) Disulfide Bonds ~<— Endo-F Treatment 4 N-Terminal Half C-Terminal Half (47,723 be, 413 (47,930 be, 431 amino acids with amino acids with eight potential one potential N-linked N-linked glycosylation sites) glycosylation site) 134 of the previous study on processing of the B antigen complex (32). It is clear that the unglycosylated precursor polypeptide, pr88, which is seen only after TM inhibition, is cotanslationally glycosylated to form gp98 (Fig. 5, lane 1). N - linked glycosylation can be blocked by TM to generate the pr88 primary precursor polypeptide (Fig. 2, compare lane 1 with lane 2; 32). gp98 is further processed to form gp100 (Fig. 5), possibly by sialylation. Then gp100 is cleaved to form gp60 and gp49 (Fig. 5; 32), which can be blocked by monensin (a monovalent ionophore inhibiting tansport of glycoprotein) (32). Based on the estimated molecular masses of approximately 44 kDa for the deglycosylated backbone polypeptides of both gp60 and gp49 on SDS-PAGE (32), and the calculated molecular masses of the predicted amino acid sequences of MDV gB homolog for the N-terminal half (413 amino acids, excluding 21 amino acids for signal peptide, based on the locations of the putative signal peptide and protease cleavage site; 26) and C-terminal half (431 amino acids) being 47,723 and 47,930 Da, respectively, the molecular masses of the amino acids for both gp60 and gp49 are approximately equal. Therefore, the size difference between gp60 and gp49 is likely due to the disproportionate extent of their glycosylation. Examination of potential N-linked glycosylation sites in the predicted amino acid sequence of MDV gB homolog (26) indicates that eight are located in the N- terminal half (excluding one site in the signal peptide) and one is located in the C-terminal half. With each glycan contributing approximately 2.5 kDa (19) to a glycoprotein’s apparent molecular weight, gp60 and gp49 would likely be derived from the N- and C—terminal halves of gp100, respectively (Fig. 6). In the MDV system a molecule called pr44 may have mistakenly been thought to be the primary precursor polypeptide of MDV gB homolog (B antigen complex) (17, 32), with a hypothesis that pr44 dimerizes to form pr88 (32), for the following three reasons. First, there was an early report that pr44 was the 135 most prominant band seen after TM teatment, using the same antisera that irnmunoprecipitated gp100, gp60, and gp49 (30). Second, several proteins (including the 44 kDa protein previously called pr44) of that approximate size were apparently nonspecifically tapped in the earlier work (17, 32); as has become readily apparent in this study during irnmunoprecipitation by the monoclonal antibody IAN86 (Fig. 2, lanes 1 to 4; the light bands at 44 kDa position were visible upon longer exposure), RaPM (Fig. 2, lanes 5 to 8), antisera against tpE-MDV-gB (Fig. 2, lanes 9 to 20) and immune chicken sera (data not shown) in this study. These nonspecifically tapped proteins seen in the irnmunoprecipitates of MDV-infected cell lysates were also observed in the irnmunoprecipitates of mock-infected cell lysates at a relatively lower level (32, unpublished result), suggesting that they may be cell proteins. Third, since gplOO of the MDV gB homolog is cleaved to form gp60 and gp49, the cleavage resulted in formation of the N- and C-terminal halves with the calculated molecular masses of their polypeptides being approximately equal, 47,723 and 47,930 Da, respectively (Fig. 6). The previous result, obtained by deglycosylation of gp60 and gp49 by endo-F and -H (32), indicated that the backbone polypeptides of gp60 and gp49 have similar estimated molecular masses (approximately 44 kDa), which are in good agreement with the calculated molecular masses for N- and C-terminal halves. Based on the above and other data, it was concluded that pr44 was the primary polypeptide, and further hypothesized that pr44 dimerizes to form pr88 (32). The result reported here indicate that the previous hypothesis is not correct and that pr88 is the primary product of the MDV gB homolog gene. The established nomenclature of this laboratory, of pr88 for the primary gene product, is used in this report for consistency with published work first establishing (l7) and then using (32) this designation. That size was based on 136 the 88 kDa estimate of both Ikuta et al. (16) and this laboratory (17). Based on their analysis of the MDV gB homolog gene, Ross et a1. (26) calculated that the unglycosylated primary gene product is 95.5 kDa, excluding the signal peptide, and in their gels a size of 94 kDa was observed. Discrepancy with the original estimate of 88 kDa (16, 17) used to establish the nomenclature adopted by this laboratory (1 7) is likely due to the variation inherent in SDS-PAGE analysis, especially between laboratories. Compounding factors could be a change (by the manufacturer) in the reported size of the phosphorylase B marker from 92.5 kDa (17) to 97 kDa (26, Fig. 2), and the use of the 116 kDa b-galactosidase marker (26); which may have resulted in a SDS-PAGE-based estimate (94 kDa) much closer to that derived from sequence data (26). Possibly the term pr88 will need to be revised, especially as new MDV nomenclature evolves to accommodate sizes more precisely determined by calculations based on open reading frames shown to encode the polypeptides (28). In that event, the term pr95.5 is suggested based on the calculated size of the predicted amino acid sequence (26), rather than the SDS-PAGE derived size (16, 17). Possibly knowledge of function and/ or virion location will be included in such new nomenclature. Since fragment 1 and 3 (their tpE fusion proteins were used to prepare antisera Z43 and Z42, respectively) are from regions located at either N- or C- terrninal parts of the putative cleavage site of the predicted MDV gB homolog amino acid sequence (Fig. 1), both Z43 and Z42 should recognize gp100, and one of the cleaved forms of gpl 00 (either gp49 or gp60), but not both of the cleaved forms, especially when the lysates were boiled. However, gp100, gp60 and gp49 were detected by both Z43 and Z42 (Fig. 4A, lanes 6 and 8, respectively). There is no significant amino acid homology found between the N- and C-terminal halves. Probably, gp49 is specifically associated with gp60 137 through a disulfide bond(s) which is(are) then dissolved on the reducing SDS- PAGE, as was the case for the two cleaved forms of EHV and VZV gB homologs (21 , 22). The fact that Z48 recognizes the undenatured gp60 and gp49 of HVT, but not those of MDV (Fig. 2, compare lane 13 with lane 15), suggest significant differences in the manner in which HVT’s epitopes are displayed compared to MDV’s. The failure of Z43, Z45 and Z42 to detect either gplOO (Fig. 2, lane 17) or the cleaved forms of gp100 (gp60 and gp49) (Fig. 2, lanes 9, 13 and 17), except upon denaturation (Fig. 4A, lanes 4, 6 and 8), is likely due to glycosylation and] or creation of secondary stuctures in the cleavage product that made certain epitopes inaccessible for recognition by antisera against tpE- MDV-gB. A similar phenomenon was reported by Whealy et al. (35); boiling of PRV-infected cell lysates increases the apparent reactivity of the mature 92 kDa protein of gII (PRV gB homolog) with 282 antiserum (raised against a denatured E. Cali-produced Cro-gII fusion protein). The conclusion that the MDV B antigen complex is encoded by MDV's homolog of the HSV gB gene will facilitate future studies on the irnmunobiology of MD. It has been known for sometime that antibodies to MDV gB homolog neutalize MDV in cell culture (16), and that afinity purified HVT gB homolog elicit partial protection against MD in chickens (23). It is not clear if this partial protection is because another antigen(s) is needed, or because of inadequate antigen, adjuvant or other factors. Identification of the gene encoding this important antigen will make it possible to use molecular biological approaches to study immunoprotection against MD in more depth, including resolution of the questions presented above. 138 ACKNOWLEDGMENTS This research was supported primarily by a contact with Intervet International B. V., Boxrneer, the Netherlands. We especially thank Dr. P. I. A Sondermeijer of Intervet, for his role as Intervet molecular virology advisor to this contact. We thank Quentin McCallum and Ruth A. Stinger for their excellent technical assistance, and Drs. Peter Brunovskis, Walter J. Esselman, Robert F. Silva, and P. I. A. Sondermeijer for their careful reviews of the manuscript and many helpful suggestions. The pATH vectors, monoclonal antibody IAN 86 and BamHI clones used in this study were kindly provided by Drs. Richard C. Schwartz, L. F. Lee, and M. Nonoyama, respectively. 10. 139 Banter, W. M., and R. A. Lasltsy. 1974. A film detection method for tritium-labeled proteins and nucleic acids in polyacrylamide gels. Eur. I. Biochem. 46:83-88. Britt. W. I., and D. Am. 1986. Synthesis and processing of the envelope gp55-116 complex of human cytomegalovinls. I. Virol. 58:185-191. Britt. W. I., and D. Auger. 1989. Processing of the gp55-116 envelope glycoprotein complex (gB) of human cytomegalovinls. I. Virol. 63:403-410. Calnek. B. K 1985. Marek's disease - a model for herpesvinis oncology. Crit. Rev. Microbiol. 12:293-320. Chen. K, P. I. A. Sondermeijer, and L. F. Velicer. 1992. Identification of a unique Marek's disease virus gene which encodes a 38 kd phosphoprotein and is expressed in both lytically infected cells and in latently infected lymphoblastoid tumor cells. I. Virol. 66:85-94. Chen. K. and L. F. Velicer. 1991. Multiple bidirectional initiations and terminations of tanscription in the Marek’s disease virus long repeat regions. I. Virol. 65:2445-2451. Chimb. R. C., and A. E. Churchill. 1969. Precipitating antibodies associated with Marek’s disease. Vet. Res. 83:4-7. Clutchill. A. B., R. C. Chubb, and W. Baxendale. 1969. The attenuation, with loss of antigenicity, of the herpes-type vinis of Marek’s disease (stain HPRS-16) on passage in cell culture. I. gen. Virol. 4:557-563. Coulson. P. M., and L. F. Velicer. 1988. Structure and complete nucleotide sequence of the Marek’s disease herpesvirus gp57-65. I. Virol. 62:2373-2379. MP.M..MR.Wilson.I'I.Camp,H.RoehLR.I.Isfort.andL.F. Velicer. 1990. Characterization of the gene encoding herpesvinis of ll. 12 13. I4. 15. 15. 17. 18. 140 turkeys gp57-65: comparison to Marek’s disease virus gp57-65 and herpes simplex virus glycoprotein C. Vinis Genes 3:211-227. WLMMalkirIsorLandY.Becker. 1988. Marek’s diseasevirus serotype-l antigens A and B and their unglycosylated precursors detected by Western blot analysis of infected cells. Virus Genes 2:5-18. thpatichD.R.,T.Zmb.M.D.Parker,S.D.L.Hurk.L.A.Babhik.and M. I. P. LawmarL 1988. Expression of bovine herpesvinls 1 glycoproteins g1 and gIII in tansfected murine cells. J. Virol. 62:4239-4248. Glaubiger, C., K Naserlarl. and L. F. Velicer. 1983. Marek’s disease herpesvinls. IV. Molecular characterization of Marek’s disease herpesvinls A antigen. I. Virol. 45:1228-1234. Glue, C. 1990. Glycoproteins encoded by Varicella-Zoster virus: biosynthesis, phosphorylation, and intacellular taficking. Annu. Rev. Microbiol. 44:59-80. Ilarta. K. S. Ueda. S. Kato. and K Hirai. 1983. Monoclonal antibodies reactive with the surface and secreted glycoproteins of Marek’s disease vinis and herpesvinls of turkeys. I. gen. Virol. 64:2597-2610. Britta. K, S. Ueda. S. Kato, and K Hirai. 1984. Processing of glycoprotein gB related to neutalization of Marek’s disease virus and herpesvinls of turkeys. Microbiol. Immunol. 28:923-933. Isfort. R. I., I. Sithole, H.-I. Eng, and L. F. Velicer. 1986. Molecular characterization of Marek’s disease herpesvinls B antigen. I. Virol. 59:411- 419. tfort. R. I., R. A. Stringer, I'I.-I. Eng, and L. F. Velicer. 1986. Synthesis, processing, and secretion of the Marek’s disease herpesvirus A antigen glycoprotein. I. Virol. 57:464-474. 19. 21. 141 Klenk. H. D., and R. Roll. 1980. Co-tanslational and post-tanslational processing of viral glycoproteins. Current Topic in Microbiol. and Immunol. 90:19-48. Konopka.I.B.,R.L.Davis.S.MWatanabe,A.S.Pontloelli.L.Schifi- Maker, N. Rosetiberg. and O. N. With. 1984. Only site-directed antibodies reactive with the highly conserved src-homologous region of the v-abl protein neutalize kinase activity. I. Virol. 51:223-232. Meridith.D.M..I.-MStocks,G.R.Whitalter,LW.Halliburton.B.W. Snowden. and R. A. Ellingtui. 1989. Identification of the gB homologues of equine herpesvirus types 1 and 4 as disulphide-linked heterodimers and their characterization using monoclonal antibodies. I. gen. Virol. 70:1161-1172. Montalvo, E. A., and C. Grose. 1987. Assembly and processing of the disulfide-linked Varicella-zoster virus glycoprotein gpII (140). I. Virol. 61 :2877-2884. Ono.K..M.Takashima.T.Ishikavva.M.Hayashi.I.Yoehida. T.Konobe,K Ikuta. K Naksjlma. S. Ueda. S. Kato, and K Hirai. 1984. Partial protection against Marek’s disease in chickens immunized with glycoproteins gB purified from turkey-herpesvinls-infected cells by affinity chromatography coupled with monoclonal antibodies. Avian Dis. 29:533-539. Payne. L. N. 1985. Marek’s Disease - Scientific Basis and Methods of Control. Martinus Nijhofi Publishing, Boston. Robblrt,A.K.D.I.Dorney,M.W.Wathen.M.E.Whealy,C.Gold.R.I. Watcn.L.E.Holland.S.D.Weed.M.Levlne,I.C.Glorloso,andL.W. Eiqutt 1987. The pseudorabies virus gII gene is closely related to the 9B glycoprotein gene of herpes simplex virus. I. Virol. 61:2691-2701. 31. 142 Ross.L.I.N.,M.Sanderson.S.D.Scott.M.MBimts,T.Doe1.andB. Milne. 1989. Nucleotide sequence and characterization of the Marek’s disease virus homologue of glycoprotein B of herpes simplex virus. I. gen. Virol. 70:1789-1804. Sarnhook. I., E. F. Frltch. and T. Maniatt. 1989. Molecular cloning: a laboratory manual. 2nd edition. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. SclntKA..G.F.deBoer.KHirai,L.F.Lee.KNakajima.R.F.Silva.and L. F. Velicer. 1988. A proposed nomenclature for the protein of Marek’s disease herpesvirus, p. 438-443. In S. Kato, T. Horiuchi, T. Mikami, and K. Hirai. (ed.), Advances in the Marek’s disease research. Iapanese association on Marek's disease, Osaka, Japan. Silva. R. F., and L. F. Lee. 1984. Monoclonal antibody-mediated irnmunoprecipitation of proteins from cells infected with Marek’s disease virus or turkey herpesvirus. Virology 136:307-320. Silva. R. F.. and L. F. Lee. 1985. Isolation and partial characterization of three glycoproteins common to Marek’s disease virus and turkey herpesvirus-infected cells, p. 101-110. In B. W. Calnek and I. L. Spencer (ed.), Proceedings of the International Symposium on Marek’s Disease. American Association of Avian Pathologist, Inc., Kennet Square, Pa. Sithole, I., P. M. Coussens, and L. F. Velicer. 1988. Identification of Marek’s disease herpesvinis B antigen’s precursor polypeptide and the gene encoding it, p. 148-154. In S. Kato, T. Horiuchi, T. Mikami, and K. Hirai. (ed.), Advances in the Marek’s disease research. Iapanese association on Marek’s disease, Osaka, Iapan. 143 Sithole, I., L. F. Lee, and L. F. Velicer. 1988. Synthesis and processing of the Marek's disease herpesvirus B antigen glycoprotein complex. I. Virol. 62:427 0-427 9. Sullivan. D. C., G. P. Allen. and D. I. O'Callagharl. 1”. Synthesis and processing of equine herpesvinis type 1 glycoprotein l4. Virology 173:638-646. Weber, K, and M. Osborn. 1969. The reliability of molecular weight determination by dodecyl-sulfate-polyacrylamide gel electophoresis. I. Biol. Chem. 244:4406-4412. Whealy,M.E.,A.KRobbins,andL.W.Enqutt. 1990.The export pathway of the pseudorabies virus gB homolog gII involves oligomer formation in the endoplasmic reticulum and protease processing in the Golgi apparatus]. Virol. 64:1946-1955. Witte, O. N., and D. F. erth. 1979. Stucture of the murine leukemia vinls envelope glycoprotein precursor. I. Virol. 29:735-743. W V Summary and Conclusions The major focus of the thesis research was to confirm our laboratory’s previous preliminary identification of the gene encoding the MDV B antigen complex (gp100, gp60, and gp49) in BamHI-H, to locate the expected duplicate copy of the gene in BamHI-D (MDV has an alphaherpesvinis genome stucture: both BamHI-D and -H contain part of long repeat regions), and to identify the HVT homolog of the MDV 8 antigen gene. Preliminary nucleotide sequence analysis suggested the presence of a spliced gene in BamHI-H, resulting in cDNA analysis. No open reading frame characteristic of a glycoprotein gene was identified in either BamHI-D or -H by DNA sequence and cDNA analyses; and no significant DNA homology was found between BamHI-D or -H and the equivalent part of HVT’s genome by Southern blot analysis. Instead, as a result of the cDNA analysis a gene encoding pp38 was identified in BamHI-H, and a gene family composed of four groups of transcript were mapped in both BarnHI-D and -H. Southern blot analysis showed that both the gene encoding pp38 and the gene family composed of four groups of tanscript are not present in HVT, suggesting that these genes are unique to MDV. The gene encoding the B antigen complex was not found in BamHI-D and -H, where it was previously thought to exist. However, identification of the gene encoding the MDV homolog of herpes simplex vims (HSV) glycoprotein B (gB) was reported by others. Therefore, determining whether the B antigen complex is the MDV gB homolog seemed to be the logical next choice in the efi'ort to identify the gene encoding the MDV B antigen complex. To determine whether the MDV gB homolog gene might encode the B antigen complex, antisera were prepared against trpE fusion proteins of the MDV gB homolog (tpE-MDV-gB). These antisera irnmunoprecipitated gp100, 144 145 gp98, gp60, gp49, and pr88. Based on size comparison, and tpE-MDV-gB competition and blocking assays, it was concluded that the MDV gB homolog is the B antigen complex and pr88 is the precursor polypeptide, while pr44, previously thought to dimerize to form pr88, is a protein nonspecifically tapped during irnmunoprecipitation. The MDV gB homolog processing was found to be consistent with those of other herpesvirus gB homologs; and appears to involve cotanslational glycosylation of pr88 to form gp98, which is further processed to yield gp100, which is then cleaved to form gp60 and gp49. The conclusion that the B antigen complex is the MDV gB homolog indicates that the previously hypothesized dimerization of pr44 to form pr88 of the B antigen complex is incorrect, clarifies the uncertainty regarding the location of the gene encoding the B antigen complex, and will facilitate future studies on the immunobiology of MD. It has been known for sometime that antibodies to MDV gB homolog neutalize MDV in cell culture and that affinity- purified HVT gB homolog elicit partial protection against MD in birds. It is not clear if this partial protection is because another antigen(s) is needed, or because of inadequate antigen, adjuvant or other factors. Identification of the gene encoding this important antigen will make it possible to use molecular biological approaches to study immunoprotection against MD in more depth, including resolution of the questions presented above. Through DNA sequence and cDNA analyses, an open reading frame (designated pp38 ORF) encoding 290 amino acids was identified in BamHI-H and is not duplicated in BamI-II-D. Using tpE-pp38 fusion proteins, antisera against pp38 were prepared. With irnmunoprecipitation and SDS-PAGE analyses, two polypeptides (38- and 24-kDa) (designated pp38 and p24, respectively) were found. No precursor-product relationship was found between pp38 and p24 by pulse-chase analysis, and only pp38 was detected by 146 irnmunoblot analysis with anti-pp38. pp38 was found to be phosphorylated, MDV serotype-l-specific, an irnmunogen in birds with MD, belong to one of the early classes of herpesvinis proteins, and it was detected in MSB-l lymphoblastoid cells in absence of IUdR teatnent. Some of the properties predicted for a protein detected by molecular methods, based on finding the ORF first, are identical to those of a 38-kDa phosphoprotein reported by others, suggesting that they are the same. However, many new pp38 properties were identified, which greatly extend our understanding of pp38. The identification of pp38 as the sole gene product of the pp38 ORF clarified the status of two other component, p24 and p41, of the previously identified group of three proteins that included pp38. They are not part of the gene product of the pp38 ORF, and they may specifically or nonspecifically associate with pp38, respectively. Detection of pp38 in the MSB- 1 lymphoblastoid cell line in absence of IUdR teatnent indicates that pp38 is constitutively expressed in the tansformed and latently infected tumor cell line, suggesting that pp38 may play a role in MDV latency and maintenance of the tansformed state of the tumor cell line. Absence of significant amino acid homology of pp38 to any known oncogenes, and expression of pp38 in attenuated nononcogenic serotype-1 MDV (reported by others) suggest that if pp38 plays a role in MDV tansfonnation, two possibilities exist: (i) it is modified during attenuation or (ii) it is necessary but insufficient and indirect, and attenuation involves another gene (3). Identification of pp38 as one of the early classes of herpesvirus proteins suggest that pp38 may have a regulatory function, as do other herpesvirus early proteins. The finding that immune chicken sera (ICS) contain antibodies against pp38 indicates that pp38 is an irnmunogen in birds with MD. Since pp38 is an irnmunogen and constitutively expressed in the MSB-l tumor cell line, and since inactivated MSB-l tumor cells 147 were found by others to induce anti-tumor immunity in birds, it is possible that pp38 may play a role in MDV tumorigenicity and may be used as recombinant DNA derived vaccine to prevent tumor formation. The gene family composed of four groups of tanscript identified with cDNA and SI nuclease analyses were mapped in the expansion regions of BarnHI-D and -H of MDV genome, and are either initiated or terminated within or near the expanded regions at multiple sites in both rightward and leftward directions. These RNAs, containing various copies of the 132-bp repeat at either their 5’ or 3’ ends, were found to be 0.67-, 1.6-, 1.8-, and 3.1-kb. Since the MDV genome expansion in the regions of BamHI-D and -H is correlated with the loss of the MDV tumorigenicity, the gene family tanscribed from the expanded regions may play a role in MDV tumorigenicity. The finding that the gene family is composed of four groups of tanscript clarifies the erroneous published data, in which the gene family was thought to be composed of two groups of exons. Because of the bidirectional tanscriptions, groups 1 and 2 tanscript are complementary to those of groups 3 and 4, respectively. Therefore, antisense regulation may be involved in the formation and functions of the gene family. DNA sequence analysis revealed that the tanscript can be initiated or terminated within the 132-bp direct repeat and contain various copies of 132-bp repeat, and each l32-bp repeat contains one TATA box and two polyadenylation consensus sequences in each direction. Taken together, these data suggest that the 132-bp repeat, and indirectly it copy number, may be involved in tanscriptional regulation and therefore in the generation of four groups of tanscript potentially responsible for MDV tumorigenicity, although this remains to be demonstated.