Transcription of gD and gI genes in BHV1-infected cells

Transcription of gD and gI genes in BHV1-infected cells

SUMIT CHOWDHURY and BHASKAR SHARMA*Biochemistry Division, IVRI, Izatnagar 243 122, India

*Corresponding author (Fax, 91-581-2303284; Email, [email protected])

Glycoprotein D (gD) and glycoprotein I (gI) genes of bovine herpes virus 1 (BHV1) are contiguous genes with 141 bpregion between the two open reading frames (ORFs). Expression of gD and gI from a bicistronic construct containingcomplete gD and gI gene has been reported either through internal ribosome entry site (IRES)-like element or throughthe scanning and leakage model (Mukhopadhyay 2000). We here show by computational and experimental means thatgD is expressed solely as bicistronic transcript comprising gD and gI coding region in BHV1-infected cells. gI ORFwas also shown to express separately. An IRES-like element was also predicted by IRES predicting software in themiddle of the gD coding region; within that region a putative promoter was also identified by promoterscan. Theintergenic region between the two ORF showed extensive secondary structure which brings the stop codon of gD veryclose to start codon of gI gene. gD gene transcript in BHV1-infected cells was solely bicistronic. gI transcript was alsopresent in the BHV1-infected cells but in low copy number. The results indicate that gI is probably transcribed from itsown transcript in BHV1-infected cells.

[Chowdhury S and Sharma B 2012 Transcription of gD and gI genes in BHV1-infected cells. J. Biosci. 37 971–977] DOI 10.1007/s12038-012-9258-7

1. Introduction

The BHV1 genome has a size of 135.3 kb and comprises twounique sequences, unique long (UL) and unique short (US)(Tikoo et al. 1995). The latter is bracketed by invertedrepeats (internal repeats and terminal repeats). During DNAreplication the UL and US regions can flip-flop relative toeach other and can generate isomeric forms (Bryan et al.1994). There are 73 open reading frames (ORFs) homolo-gous to similar genes in other alphaherpesvirus in the BHV1genome (Wirth et al. 1989). Many of these genes are over-lapping, co-terminal at the 3′-end and use different readingframes for expressing respective proteins (BHV1 genome,Ac. No. NC001847). gD and gI genes are not overlappingbut are contiguous with a 141 bp region between them(BHV1 genome, Ac. No. NC001847).

gD and gI are surface glycoproteins of the BHV1 virus.gD (US6) with 417 amino acids (MW 71 kDa) is a majorglycoprotein of the virus envelope and is essential for virusentry and cell-to-cell spread (Whitbeck et al. 1996). gI (US7)is a minor glycoprotein of BHV1 and has an N-terminalextracellular domain and C-terminal cytoplasmic domain.The extracellular domain has cysteine-rich domain that is

also found in homologous proteins of other alphaherpesvi-ruses (Petrovskis et al. 1986). gI is important for the viru-lence of the virus but is not important for virus replication incell culture (Jacobs et al. 1994).

While searching for gD transcripts from a cDNA libraryof BHV1, we found mostly bicistronic transcripts and a fewtricistronic transcripts. No monocistronic gD gene transcriptswere found (Sharma B, unpublished data). A search for thetranscription stop signal between gD and gI indicated itsabsence in this region. In a Master’s thesis (Mukhopadhyay2000) it was reported that from the bicistronic gD–gI con-struct, both gD and gI are expressed.

In eukaryotes, in general, the translation initiationbegins from the first AUG by the scanning mechanism.In this mechanism, the 40S ribosomal unit binds to thecap structure at the 5′-end of mRNA and scans mRNAfor the initiation codon AUG. The assembly of ribosom-al takes place at this AUG and translation begins. At thestop codon the ribosomal disassemble and further trans-lation does not occur. In this mechanism, translationfrom downstream AUG does not take place normally,but in exceptional cases initiation may take place fromthe downstream AUG (Pestova et al. 2007). Thus, in

http://www.ias.ac.in/jbiosci J. Biosci. 37(6), December 2012, 971–977, * Indian Academy of Sciences 971

Keywords. BHV1; bicistronic; glycoprotein D (gD); glycoprotein I (gI); transcription

Published online: 28 September 2012

eukaryotes, generally as a rule of thumb, monocistronicmRNAs are produced and translated. In many viruses, poly-cistronic mRNAs are produced and proteins expressedthrough them (Samuel 1989; Kang et al. 2009). The expres-sion may be through leaky scanning mechanism or by reini-tiation of translation from downstream AUG (Kozak 2001)or through internal ribosome entry site (IRES). IRES ele-ments were first identified in RNA viruses. Kozak (2003)has challenged the concept of IRES based on the methodadopted for studying IRES activity in cells. It was proposedthat the IRES activity in transiently infected cells may resultfrom the presence of cryptic promoters within the DNAconstruct itself (Kozak 2003). The first report of IRES inDNA virus came from herpes virus (Bieleski and Talbot2001; Grundhoff and Ganem 2001; Low et al. 2001). Sincethen few more IRESs have been identified in DNA viralgenomes, viz herpes simplex virus (HSV; Griffiths andCoen 2005), simian virus 40 (SV40; Yu and Alwine 2006)and white spot syndrome virus (WSSV; Han and Zhan2006). In many cases it has been established that claimedIRES activity was because of the presence of a crypticpromoter activity (Bert et al. 2006).

The purpose of this study was to establish gD and gI geneexpression in BHV1-infected cells, and to check whether gDis expressed only as bicistronic transcript and also to check ifgI transcript is also as expressed as monocistronic transcriptin BHV1-infected cells.

2. Material and methods

2.1 Cells and virus

MDBK cells were used for growing the BHV1 virus and forstudying the transcript number in virus infected cells. Verocells were used for studying reporter gene activity. BHV1virus used in this study was obtained from the CADRADdivision of this institute. The cells were grown in DMEMsupplemented with 10% FBS.

2.2 Primers, vectors and constructs

The sequence and location of the primer used in this study isgiven in table 1. For all amplification, proofreading enzymepfu was used in this study.

pDsRed N1 express vector (Clonetech) was digested withVspI and NheI. This removes CMV promoter enhancerregion. Both the promoter and vector were purified andblunt-ended, religated and cloned. The cloned vectors werescreened for pDsRed vector with CMV in reverse orientationby SnaB1 and EcoRI digestion. The digestion with theseenzymes results in different size fragments for CMV inoriginal and reverse orientation. The vector with CMV-

Enhancer region in reverse orientation (pDsRed CMV Rev)was used for studying the promoter expression. The differentpromoter regions were amplified by PCR and cloned intoXhoI and EcoRI site of this vector. The original vector wasused as positive control and the vector with promoter inreverse orientation and no insert was used as negativecontrol.

2.3 Promoter activity

The putative promoter region was identified by promoter-scan (Prestridge1995). Seven different primer pairs weredesigned (table 1) to amplify different regions containingthe putative promoter sequence except one. These primershad XhoI and EcoRI sites. After amplification the amplifiedproduct was digested with these enzymes and cloned intoXhoI and EcoRI sites of pDsRedCMVRev vector. The pro-moter activity was studied by flow cytometry.

The following protocol was adopted for studying pro-moter activity. Cells (Vero) were grown in 6-well platesto 70% confluency in DMEM supplemented with 10%FBS. The cells were transfected with different promoterconstructs using Escort transfection reagent (Sigma) asper manufacturer’s recommended protocol. The expres-sion was studied at 48 h post transfection. At the ap-propriate time (48 h post transfection), media wasdiscarded, the cells detached with trypsin, washed withPBS and resuspended in PBS. The fluorescence of re-porter gene was measured in a flow cytometer (FACS-CALIBUR, BD).

2.4 Computational analysis

The RNA secondary structure in the intergenic region waspredicted by Italian RNA fold server of Polytechnic Univer-sity of MARCHE ANCONA, Italy (Markham and Zuker2005). IRES was identified by a Web-based tool IRSS whichidentifies IRES elements based on minimum free energystructure prediction and comparative sequence analysis(Wu et al. 2009). The complete gD and gI gene sequence(118896–121436, Ac. No. NC001847 ) was submitted forIRES identification. This region was also screened initiallyfor promoters with promoterscan (Prestridge 1995) and sub-sequently with BDGP server at Berkley (Reese 2001).

2.5 Copy number of gD, gD–gI and gI transcript

qPCR was done to calculate copy number of gD and gItranscript. Primers were designed from the coding region ofgD and gI gene. For estimating the copy number of gD–gIbicistronic transcript, primers from gD and gI coding regionswere used (Primer sequence and position, table 1). PCR was

972 Sumit Chowdhury and Bhaskar Sharma

J. Biosci. 37(6), December 2012

performed after extracting total RNA from MDBK cellsinfected with BHV1 virus at 18 h post infection. RNAwas extracted with trizol and treated with Dnase (Rnasefree), phenol extracted, reprecipitated and dissolved inRnase-free water; cDNA was made and used in qPCR.

Absence of DNA in the RNA sample was checked byrunning PCR with BHV1 primers. A plasmid containingfull-length gD–gI transcript was used to run standardcurve. The real-time PCR machine used was iCycler iQTM5(Biorad).

Table 1. Sequence and position of the primers used in this study

Sl. No.PCR productspecification

Primerspecification Primer sequences

Position inBHV-1 genome

PCR productsize (bp)

1 Clone A Forward 5′GAGCTCGAGGCAGCCGGGCGGGAG 3′ 119773-119798 163Reverse 5′TGCGTGATGAATTCGAG 3′ 119919-119935

2 Clone B Forward 5′AAACTCGAGTTCTGCGCCTGACGTATCTCA 3′ 119560-119586 376Reverse 5′TGCGTGATGAATTCGAG 3′ 119919-119935

3 Clone C Forward 5′GGTCTCGAGGCCGGCCGGGGAC 3′ 119462-119483 474Reverse 5′TGCGTGATGAATTCGAG 3′ 119919-119935

4 Clone D Forward 5′ TCTCGAGGAGTCGAAGGGC 3′ 119666-119684 135Reverse 5′ CGGAATTCCCCTCCCGCCCG 3′ 119780-119800

5 Clone E Forward 5′ TCTCGAGGAGTCGAAGGGC 3′ 119666-119684 293Reverse 5′CGAATTCGGGGCGGGCGGGGGGTGC 3′ 119933-119958

6 Clone F Forward 5′ TCTCGAGGAGTCGAAGGGC 3′ 119666-119684 509

Reverse 5′GGGAATTCGGGGGAGGGCCTAG 3′ 120152-120174

7 Clone G Forward 5′ TCTCGAGGAGTCGAAGGGC 3′ 119666-119684 646Reverse 5′ CCAGAATTCAGAGCAACAGG 3′ 120291-120311

8 Real time PCRfor gD gene

Forward 5′ CCCGATGCCGCGATACAACT 3′ 119000-119019 240Reverse 5′ CGCACCCGCTCTCGATCTTG 3′ 119219-119238

9 Real time PCRfor gI gene

Forward 5′GCGGCGCTACAACGGGACC 3′ 120476-120494 208Reverse 5′ GGCGATGGAGAAGAGCACGTG 3′ 120660-120683

10 5′ RLM- RACE Outer 5′GCGAGCACAGAATTAATACGACT 3′ 5′RACE Adapter 500

Inner 5′CGCGGATCCGAATTAATACGACTCACTATAGG 3′ 5′RACE Adapter

11 gD-gI Intergenicregion

Forward 5′ GAACTGCAGTCGCGAGCGCGCCGGAGGA3′ 120130-120157 158Reverse 5′GGGGGATCCTGCCCGGGTGAGCGGCCTA 3′ 120258-120287

Figure 1. The secondary structure of region between gD stop codon and gI start codon (position 120147–120289).

Transcription of gD and gI genes in BHV1-infected cells 973


2.6 Transcription start site

Transcription start site of gI was identified by using anRNA-ligase-mediated rapid amplification of cDNA ends(RLM-RACE) kit essentially as described by the supplier(Invitrogen). RNA was extracted from virus-infectedMDBK cells at 18 h post infection. RLM product wascloned into pGEMT Easy TA vector and was sequencedthrough vendors (Ocimum Biosolutions, Hyderabad).The sequence was aligned with BHV1 sequence andthe TSS identified from the aligned sequence.

3. Results

3.1 Computational analysis

The 137 bp intergenic region between the gD and gI codingregion showed extensive secondary structure, which bringsthe stop codon of gD gene very close to start codon of gIgene (figure 1). A search for the putative promoter withpromoterscan (Prestridge 1995) revealed a promoter-likesequence in the gD region at position 119907–119927; basedon this result, different regions upstream and downstreamwere cloned for studying promoter activity. Subsequently,

the BDGP server at Berkley identified four promoter-likeregions (table 2) in the same sequence. A search for IRES inthe intergenic region through an online search engine forIRES did not show IRES activity in the intergenic region(Wu et al. 2009). When the complete gD–gI sequence in-cluding the intergenic region was submitted for IRES, aputative IRES region was identified in the gD coding regionfrom 119793 to 120091.

3.2 gD transcript is polycistronic

Real-time PCR with primers spanning gD, gI or gD–gIcoding regions show (figures 2 and 3) that the copy numbersof gD and gD–gI were almost the same, thus indicating thatall the gD transcript is expressed as bicistronic gD–gI. Thecopy number of gI was slightly higher than gD and gD–gIcopy numbers. The copy number of gI includes gD–gItranscript also. The actual copy number of monocistronicgI is very low, about 30,000 to 40,000 copies versus about1 million for gD–gI transcript (gI transcript minus gD–gItranscript). The results are compatible with the fact that gD isa major and gI is a minor glycoprotein of BHV1 (Tikoo et al.1995). The experiment was repeated twice. Similar resultswere obtained in the different experiments. The data fromone experiment has been shown.

Table 2. Promoter sequence identified by BDGP server at Berkley

Start End Score Promoter Sequence

1150 1200 0.94 gcgcgtacttcgtctatacgcgccggcgcggtgcgggtccgctgcccaga1194 1244 0.99 cccagaaagccaaaaaagctgccggcctttggcaacgtcaactacagcgc1309 1359 0.89 ccgagccgcgcggggcgggagataaagcgcccgcgcgtcggcgactcaag1319 1369 1.00 cggggcgggagataaagcgcccgcgcgtcggcgactcaagccattgccgc

The sequence with enlarged base indicates putative TSS.

Figure 2. gD and gI copy number in virus infected cells. 1, 2 and 3 represents results from three different experiments.



Figure 3. gD and gD-gI copy number in BHV1-virus-infected MDBK cells. 1, 2 and 3 represents results from three different experiments.

Figure 4. (A) Relative positions of BHV-1 gD and gI gene and different reporter constructs. IR is the region between gD and gI ORFs.The position of putative promoter predicted by promoter scan was between positions 119907–119927 and the position of IRES elementbetween 119793 and 120091. The position of TSS of gI is at 120254. The upward and downward directed arrows indicate start and stopcodons. (B) Relative expression of reporter gene (geometric mean, X axis) from different promoter constructs (Y axis).



3.3 Promoter activity in gD coding regionand transcription start site for gI gene

Different upstream and downstream regions from the puta-tive promoter identified by promoterscan expressed the re-porter gene to slightly different extents (figure 4).

RLM-RACE identified only one TSS for the gI gene atposition 120254 about 35 bp upstream from the translationinitiation site (figure 4). The promoter identified by promo-terscan was about 328 bp upstream from the TSS identifiedby RLM-RACE. The BDGP server predicted four promoter-like sequences in the region upstream of the gI start codon.Two of these promoter-like sequences were in the 137 bpintergenic region. The promoter with the highest predictedscore was immediately upstream of TSS. The TSS identifiedby the BDGP server for this promoter at position 120252,just two bases upstream from that identified experimentallyby RLM-RACE. From the results it appears that the regionjust upstream of the TSS identified by RLM-RACE is thepromoter for the gI gene, although most of the region in thegD coding region has promoter-like activity, as seen exper-imentally and through promoter prediction software.

4. Discussion

This work was initiated primarily to validate our assumptionthat the gD gene transcript in BHV1-infected cells is abicistronic transcript which expressed both gD and gI geneseither through an IRES-like element or through the scanningand leakage model. Search for IRES-like element through insilico approach indicated an IRES-like element near themiddle of gD coding region, which overlaps a putativepromoter, making it difficult to study IRES activity of thisregion. We cloned different regions containing promoterregions to check promoter-like activity. These regions hadshown promoter activity, thus suggesting that gI is gettingtranscribed from its own promoter. We performed Northernblot to check for gI transcript. The data was however incon-clusive (data not shown). The intergenic region between gDand gI showed extensive secondary structure, which bringsthe stop codon of gD gene very close to start codon of gIgene, suggesting an intriguing possibility that, instead ofdisassembling at stop codon, the ribosome may just moveover and initiate expression of gI gene (figure 1). To checkthis possibility we had made a construct in which gD–gIintergenic region was cloned between two reporter genesDsRed and GFP, but we could not find any detectableexpression of the second reporter gene (data not shown).Real-time PCR data indicated that gD gene is transcribedonly as gD–gI bicistronic transcript in BHV1-infected cells.Real-time PCR data did indicate the existence of gI transcriptin BHV1-infected cells. To conclusively prove that gI tran-script is present in BHV1-infected cells, we looked for its

transcription start site. There was one single transcriptionstart site for gI. This transcription start site was in the inter-genic region about 328 bp downstream from the putativepromoter site. We did extensive literature search but couldnot find one instance where the transcription start site islocated at such a distance from the putative promoter site.Subsequently we searched for promoter with BDGP server atBerkeley (Reese 2001). BDGP was originally developed forpredicting Drosophila promoters but can predict eukaryoticpromoters also. This program identified four promoter-likesequences very close to each other just upstream of gI TSSwith predicted TSS was just 2 bp from the one we identifiedexperimentally. Interestingly, none of the promoter-likesequence predicted by DBGP server was near that predictedby promoterscan. We presume that the promoter closest toTSS is the one used by BHV1 virus for transcribing gI gene.The assumption that expression of gD and gI gene productfrom full-length gD and gI bicistronic construct is eitherthrough an IRES-like element or by the scanning and leak-age model (Mukhopadhyay 2000) is not tenable from ourresults. It has been argued by Kozak (2003) that in mostcases, expression through IRES is through cryptic promotersand the low amount of transcript can be missed in Northernblot. Our Northern blot results were not conclusive; how-ever, we did find TSS for gI and gI transcript by real-timePCR. Based on these observations we are of the opinion thatfrom an gD–gI bicistronic construct, gI gene is transcribedby its own promoter.

Expression of BHV-1 gD gene as bicistronic transcript inBHV1-infected cells apparently does not confer any advan-tage to the virus and the reason for its bicistronic expressionis yet to be explored.

Acknowledgements

The authors thank ICAR/IVRI for financial support for con-ducting this research.

References

Bert AG, Grepin R, Vadas MA and Goodall GJ 2006 AssessingIRES activity in the HIF-1 and other cellular 5′ UTRs. RNA 121074–1083

Bieleski L and Talbot SJ 2001 Kaposi's sarcoma-associated herpes-virus vcyclin open reading frame contains an internal ribosomeentry site. J. Virol. 75 1864–1869

Bryan LA, Fento RA, Misr V and Haines DM 1994 Fatal generalizedbovine herpes virus type 1 infection associated with a modified-live infectious bovine rhinotracheitis parainfluenza-3 vaccine ad-ministered to neonatal calves. Can. Vet. J. 35 223–228

Griffiths A and Coen DM 2005 An unusual internal ribosome entrysite in the herpes simplex virus thymidine kinase gene. PNAS102 9667–9672



Grundhoff A and Ganem D 2001 Mechanisms governing expres-sion of the v-FLIP gene of Kaposi's sarcoma-associated herpes-virus. J. Virol. 75 1857–1863

Han F and Zhan X 2006 Internal initiation of mRNA translation ininsect cell mediated by an internal ribosome entry site (IRES)from shrimp white spot syndrome virus (WSSV). Biochem.Biophys. Res. Commun. 344 893–899

Jacobs L, Mulder WAM, Priem J, Pol JMA and Kimman TG1994 Glycoprotein I of pseudorabies virus (Aujesky’s dis-ease virus) determines virulence and facilitates penetrationof virus into central nervous system of pigs. Acta Vet. Hung.42 289–300

Kang S, Leu J, Wang H, Chen L, Kou G and Lo C 2009 Polycis-tronic mRNAs and internal ribosome entry site elements (IRES)are widely used by white spot syndrome virus (WSSV) struc-tural protein genes. Virology 387 353–363

Kozak M 2001 New ways of initiating translation in eukaryotes?Mol. Cell. Biol. 21 1899–1907

Kozak M 2003 Alternative ways to think about mRNA sequencesand proteins that appear to promote internal initiation of trans-lation. Gene 318 1–23

Low W, Harries M, Ye H, Du M, Boshoff C and Collins M 2001Internal ribosome entry site regulates translation of Kaposi'ssarcoma-associated herpesvirus FLICE inhibitory protein. J.Virol.75 2938–2945

Markham NR and Zuker M 2005 DINAMelt web server for nucleicacid melting prediction. Nucleic Acids Res. 33 W577–W581

Mukhopadhyay B 2000 Cloning and expression of bovine herpesvirus 1 glycoprotein 1, MVSc thesis, Indian Veterinary ResearchInstitute, Izatnagar

Pestova TV, Christopher JRL and Hellen UT 2007 The Mechanismof translation initiation in eukaryotes in translational control inbiology and medicine. CSH Monogr. 48 87–128

Petrovskis EA, Timmins JG and Post LE 1986 Use of lambda gt11to isolate genes for two pseudorabies virus glycoproteins withhomology to herpes simplex virus and varicella-zoster virusglycoproteins. J. Virol. 60 185–193

Prestridge DS 1995 Predicting Pol II promoter sequences usingtranscription factor binding sites. J. Mol. Biol. 249 923–932

Reese MG 2001 Application of a time-delay neural network topromoter annotation in the Drosophila melanogaster genome.Comput. Chem. 26 51

Samuel CE 1989 Polycistronic animal virus mRNAs. Prog. NucleicAcid Res. Mol. Biol. 37 127–153

Tikoo SK, Campos M and Babiuk LA 1995 Bovine herpesvirus 1(BHV-1: biology, pathogenesis, and control). Adv. Virus Res. 45191–223

Yu Y and Alwine JC 2006 19S late mRNAs of Simian Virus 40have an Internal Ribosome Entry Site upstream of the virionstructural protein 3 coding sequence. J. Virol. 80 6553–6558

Wirth UV, Gunkel K and Schwyzer M 1989 Spatial and temporaldistribution of bovine herpes virus transcripts. J. Virol. 63 1465–1468

Whitbeck JC, Knapp AC, Enquist LW, Larence WC and Bello LJ1996 Synthesis, processing and oligomerization of ovine herpesvirus 1 gE and gI membrane proteins. J. Virol. 70 7878–7884

Wu TY, Hsieh CC, Hong JJ, Chen CY and Tsai YS 2009 IRSS: aweb-based tool for automatic layout and analysis of IRES sec-ondary structure prediction and searching system in silico. BMCBioinformatics 10 160

MS received 18 May 2012; accepted 02 August 2012

Corresponding editor: SHAHID JAMEEL



Documents

Transcription of gD and gI genes in BHV1-infected cells