. 7
( 61 .)


smallest posstble gel volume. If the correct bands cannot be seen, then the annealmg
temperature (50-65”(Z) and the MgCl, concentration (0 5-5 mA4) should be optimized
5. Setup a secondPCR reaction mtx contammg300 ng each of primers p 1 and p2
and 3 pL of each of the two gel shces (first melt at 7O”C, 5 mm, and mix), and
cycle as before, wtth optrmtzatron if necessary
6. Clone the products. Make up the PCR reaction to 400 uL with water, phenol-
extract, and prectpttate the DNA Cut wtth the appropriate restrtction enzyme m
100 pL for 1 h at 37”C, phenol-extract and prectpitate the DNA, and run the
entire reaction on an LMP gel Excise the appropriate band and hgate to suitably
prepared vector m the gel as m Section 3 1.1
HS V Mutagenesis 41
.. . ..........................
..... ..v ...... .... .,---o-o---) .....
.. . ....... ............. -.
......... - ..

primers for pomt mutation

.L7z+ ..\I,
e.g. 5 base insertion requires at least 15 bases on each side.

primers for small insertion

during PCR 1

non-complementary gene-specific
port˜ons+complementafy Insert
portions (at least 15 bases) allows
ampltficatron during PCRs 1 and 2
to Include ˜15 base Insertron

after annealing of products
.4-------i A and B

:™ overlap region during PCR 2

primers for larae msertron

20 bases
20 bases
allows amplification
allows amplification
during PCR 2
during PCR 1
\ ,, .” ,.,


deleted region

Fig. 2. The design of gene-specific primers for mutagenesis by overlap extension.

Alternatively, PCR products can be cloned directly using, for example, the T-
vector system from Promega.
7. Check for the mutation by appropriate restriction digestion, and finally by sequencing.

1 Sambrook, J., Frttsch, E F , and Mamatrs, T (1989) Molecular Clontng A Labora-
tory Manual, 2nd ed., Cold Sprmg Harbor Laboratory, Cold Spring Harbor, NY
2 McKmght, S L and Kingsbury, R. (1982) Transcriptional control srgnals of a
eukaryottc protem codmg gene Sczence 217,3 16-324
3 De Wind, N , ZiJdervtld, A , Glazenburg, K., Gielkens, A , and Berns, A (1990)
Lurker msertron mutagenesrs of herpesvnuses mactrvatron of smgle genes wtthrn
the Us region of pseudorabtes v˜ruusJ Virol 64,469 l-4696
4 Parker, R C. (1980) Conversion of ctrcular DNA to lmear strands for mapping
Methods EnzymoI 65,4 15426
5 Kunkel, T. A (1985) Rapid and effictent sue-specific mutagenesrs without phe-
notyptc selectton Proc Natl Acad Scl USA S&488-492
6 Deng, W P and Ntckeloff, J A. (1992) Sate directed mutagenests of vrrtually any
plasmrd by elimmatmg a unique sue. Anal Bzochem 200, 8 l-88
7 Ho, S N., Hunt, H. D , Horton, R. M , Pullen, J K , and Pease, L R. (1989) Stte-
directed mutagenesrs by overlap extension usmg the polymerase cham reaction
Gene 77,5 1-59
Saturating Mutagenesis and Characterization
of a Herpesvirus Genome Using In Vivo Reconstitution
of Virus from Cloned Subgenomic Regions
Niels de Wind, Maddy van Zijl, and Anton Berns

1. Introduction
The study of genome structure and gene function is pivotal in understandmg
the mechanisms of rephcation, pathogenesis, and virulence of herpesviruses.
In this respect, mutagenesis and sequence analysis of genes encoded by the
vrrus are of great importance. However, the herpesvirus genomes are large,
with sizes ranging between 120 and over 200 kbp and encoding between 70
and 200 genes (see ref. 1 for a review). This large size hampers handling and
systematic mutagenesis of the vu-us genome usmg standard modern molecular
biology techniques. Most current methods of mutagenesis therefore do not rely
on direct modificatron of the viral genome m vitro but depend on exchange in
viva, by homologous recombmation, of a viral gene by a copy of the latter gene
that is truncated in vitro by insertion of a marker gene. Mutant vuus progeny
can be screened or selected for, depending on the marker gene that is used.
Commonly used marker genes are thymidine kmase and lacZ. This proce-
dure is generally used, reliable, and has yielded a wealth of mformation on
the function of herpers simplex virus type 1 (HSV-1) encoded genes.
However, it requires prior mapping and cloning of every gene to be mutag-
enized and is therefore less feasible if the virus is a novel or less-well-known
One such less-well-characterized herpesvirus is the porcine alpha-herpesvirus
pseudorabies virus (PRV, synonyms: suid herpesvirus type 1, Auleszky™s dis-
ease vu-us). To explore the contents of the virus genome and to study gene
function, we have developed a method to mutagemze the virus m vitro without
prior knowledge of the gene structure. Here, we describe how to:
From Methods fn Molecular Medme, Vol 10 Herpes Smplex Vrus Protoco/s
Edlted by S M Brown and A R MacLean Humana Press Inc , Totowa, NJ

de Wind. van ZijL and Berns

Subcloning™of the virus Reconstitution of™the virus genome b>
A genome as large overlapping cotransfection of cloned fragments
fragments (overlap recombination)

Insertion. at a random site, of an oligonucleotide, containing stop- codons in all
reading frames as well as a unique restriction site C7)

Subcloning of small fragments flanking the inserted mutagenic
c oligonucleotide in each mutant
dX aX db

-Cross hybridisation studies
-Transcript mapping
-Sequence analysis
I, Make the genome of a herpsirus amenable to modification in Litro:
2. Gencratc a large series of\ irus mutants, lbith cnch mutant carrying zm inserion of
an oligonuclcotidc in a single gene. abolishing the exprccsion of that gene: and
3. llse these mutants to obtain gene-spcctfIc probes to explore the gene layout of
the \ irus.

The method, as outlined in Fig. 1, relies on cotransfcction of four to five
cloned o\crlapping DNA fragments that together constitute the cntirc virus
Characterization of a Herpesvms Genome

genome. In vivo, replicatmg virus 1sgenerated efficiently and faithfully after
recombination between the homologous ends of the DNA molecules (“overlap
recombination,” Fig. 1A). Mutagenesis of these individual clones is performed
by insertion of an oltgonucleottde that contains stop codons m all reading
frames and a unique restriction site, at a random location, m such a large cloned
virus fragment (Fig. 1B). This 1sfollowed by reconstltutton of mutant virus by
overlap recombmatton. To obtain mformatton on the identity of mutagenized
genes, small DNA fragments flankmg the inserted ohgonucleotide in every
mutant are subcloned. These subclones are subsequently used for cross-
hybrtdizatton studies with a prototype herpesvn-us, for transcript mapping and
for direct sequence analysts (Ftg. IC).
We ˜111describe the background and rationale of the different steps of the
procedure outlined in this Section. In Sections 2. and 3., we provide a detailed
description of the techniques mvolved. In addition, in Section 4., we will pro-
vide the reader with supplementary information and hints on the techniques
1.1. The Generation of Large Overlapping Clones
of a Herpesvirus Genome for Overlap Recombination
In 1987 we demonstrated that cotransfection of a PRV genome fragment
lacking the Unique Short (US) region, together with a cloned fragment con-
taining a mutagemzed US region, resulted in the regeneration of rephcatlon

Fig 1 (opposzte) Outline of the procedures described m this Chapter Top: the
genome of PRV consisting of a Umque Long (II,) and a Unique Short (Us) region, the
latter being bracketedby the Internal Repeat(IR) andthe Terminal Repeat(TR) Small
verttcal bars indicate the cleavage sites of a (hypothetical) diagnostic enzyme. (A) The
virus genomic mformation is cloned as a set of four to five cosmids, each cosmid
containing an adJacent but overlapping region of the virus genome.Cotransfection of
the insert fragments leads to the regeneration of intact virus genome and progeny vu-us
(“overlap recombmation”) (B) An individual cosmrd clone IS used for mutagenesrs by
insertion, at a random site, of an ohgonucleotide that contams stopcodons m all read-
ing frames, abrogating translation, and also a restriction site that is normally absent
from the cosmid. In this way, a large series of mutant cosmid derivatives IS generated,
each mutant cosmid derivative bearing the oligonucleotide at a single and umque site
Mutant virus strains, each containing the oligonucleotide at a single and umque site,
are subsequently generated by overlap recombmations. (C) Of every mutant cosmid
derivative, small genomic fragmentsare subcloned,that flank the insertedohgonucle-
otide at both sides. These “tags” can be used to explore the virus genome, as probes for
crosshybrtdization studies with a prototype herpesvirus, for the mappmg of vu-us-
encoded transcripts, and for direct DNA sequencmg to identify the gene in which the
oligonucleotide IS inserted in
de Wmd, van Z# and Berns
competent vu-us (2). Apparently, ligation in viva between the ends of both DNA
molecules is responsible for this (see Note 1). This experiment did show that the
virus could efficiently be reconstituted from large subgenomic fragments of which
(at least) one was cloned and mutagemzed m vitro. Since the wild-type US frag-
ment was absent m the transfection, screening or selection for the mtroduction of
the mutation was avoided. Besides the gam in time, regeneration of virus by
cotransfection of large, excluding, genomic DNA fragments enhances both the
subtlety and the versatthty of mutations to be introduced mto the vnus genome.
It is well-established that, m cultured cells, extrachromosomal recom-
bination between homologous DNA stretches is a rapid and efficient process,
provided that the homology is present at the end of the lmearized DNA mol-
ecules. In the early 1980s this property of the cell already was used to recon-
stttute viable adenovnus after cotransfection of two overlappmg (linearized)
cloned fragments of the virus genome (3,4). Recombmatron m viva also under-
lies most established methods to mutagemze the HSV- 1 genome, as described
above (see Chapter 4).
These results led us to mvestigate whether cotransfectlon of multiple,
overlappmg, cloned fragments, covering the entire PRV genome, resulted m
the efficient reconstitution of viable virus via homologous recombmation
between the homologous ends of the lmearized virus fragments (2) To this
end, a set of cosmids was constructed, each cosmid contannng an adlacent but
overlappmg region of the vnus genome The clonmg procedure is depicted m
Fig. 2. Briefly, purified PRV genomtc DNA was mechanically sheared and
subgenomic fragments with sizes between 35 and 45 kbp were isolated and
cloned m a cosmid vector, packaged in phage lambda and used to infect
Eschevichza coli The resultmg subgenomic clones of the vrrus were accurately
mapped by restriction digest mapping. A set of three clones, overlappmg at the
ends by 0.5 and 1.5 kbp (together comprismg the Unique Long [UL] and a
copy of the Internal Repeat [IR] of the virus), was selected. In this specific
case, a fourth overlappmg fragment, containing the US region, was (for techm-
cal reasons) not cloned as a cosmtd, but cloned separately as a 2%kbp Hind111
fragment (the overlap here is about 10 kbp). These four overlapping fragments
were purified to remove vector sequences and cotransfected mto permissive
cells (see Note 2). This resulted in the highly efficient production of PRV.
Even cotransfection of another set of five, instead of four, overlappmg frag-

Fig. 2 (opposrte)Outline for the procedure for cloning the virus genomeas a a set
of four to five cosmids,eachcosmldcontaining an adjacentbut overlapping region of
the genome. The small black circle representsthe cosmid vector c. cossite, required
for packaging of the large DNA fragments in phagelambda heads Details of the pro-
cedure are explained in the text.
“L, , I
I” ” I
Shearing of the virus genome followed by size selection, Cleavage, blunting
blunting and linker addition and linker addition


. 7
( 61 .)