gent should be low enough not to have any signtficant influence on the charge of the
protein. An example of the effect of SDS ISshown in Fig. 1.The spotmarked with an
arrowhead is almost undetectablem the absenceof SDS.
One should keep in mmd that the solubthty of a protein may change from
the state immediately after translation to its final modtficatton and locahza-
tion. Thus, the spot labeled VP1 l/12 in Fig. 2, which is supposed to be the
Analysis of HSV Polypeptides 99
x 10 -3
Fig. 1. Effect of SDS on solubilization of proteins for 2D gel electrophoresis. BHK cells
were infected with HSV-2 and proteins labeled with [â€œ?S]-methionine. One portion of the
proteins was solubilized in 9.5M urea, 2% NP-40 and 5% mercaptoethanol as described in
Section 3.7. Another portion was solubilized in the same mixture except that SDS was
added to a final concentration of 1.3%. Undissolved material wasremoved by centrifugation
for 1h at 156,OOOgin a Beckman SW50.1 rotor. The slab gel contained 5-12.5% polyactyla-
mide crosslinked with BIS. The arrowhead showsa protein solubilized by SDS only.
tegument protein VP1 I/ 12 (U, 48) is quite substantial in the gel after in vitro
translation, more faint in material from infected cells, and almost undetectable
at the expected position in material from virions and light particles. This disap-
pearance is related to an increased amount of protein unable to enter the pH
gradient gel, as indicated by the arrowhead.
The solubilization procedures mentioned earlier result in denaturation of the
proteins. For certain purposes, however, it may be crucial to preserve their
native forms. A nondenaturing system for polyacrylamide gel electrophoresis
has been used successfully to localize discontinuous epitopes of glycoprotein
gD (24). In this system the concentration of SDS was reduced to 0.1% as com-
pared to 2% in ordinary SDS-PAGE.
2.1.3. Artifacts from the Reagents and influence
of the Crosslinkers on the Apparent M,
Urea used for solubilization and electrophoresis in the pH gradient may be
degraded to form cyanate ions that react with the proteins and introduce arti-
factual charges. One way of avoiding this is to keep the sample on dry ice
during all handlings, except when dissolving and loading it on the gel.
Silver staining of the slab gel may result in horizontal streaking that is usu-
ally caused by mercaptoethanol. It can be reduced by filtering the reagent. Some
researchers have successfully substituted mercaptoethanol with dithiothreitol
Haarr and Lange/and
Fig. 2. Variation of the solubility of a protein. Proteins labeled with [sSS]-methion-
ine were solubilized in the presence of urea, NP-40, and mercaptoethanol as described
in Section 3.7., and then subjected to 2D gel electrophoresis. Gradient gels of 5512.5%
polyacrylamide crosslinked with BIS were used. Upper panel: RNA was isolated from
BHK cells infected with HSV-1 and used for in vitro translation as described in Sec-
tion 3.6. Middle panel: Proteins in extract from BHK cells infected with HSV-1. Lower
panel: HSV-1 virions were radioactively labeled and purified in a Nycodenz gradient
as described in Section 3.3. Some HSV- 1 specific proteins are indicated. The relative
amount of VP1 1112 migrating into the pH gradient gel is different in the three panels
and supposed to reflect different solubility.
and observed that 65 mM is the optimal concentration (2.5). Vertical
streaking after silver staining may be caused by dust present on the glass plates
during casting of the slab gel.
Separation on 2D gels demonstrate clearly that the crosslinkers in the slab
gel affect the migration of some proteins. Figure 3 shows analysis of the same
Analysis of HSV Polypeptides 101
Fig. 3. Effects of the crosslinkers on the relative migration of proteins in polyacry-
lamide gels. Human fetal lung (HFL) cells were infected with HSV-1, proteins labeled
with [35S]-methionine and subjected to 2D gel electrophoresis. Two samples were run
identically in the first dimension, then either in a 7.5% polyacrylamide gel crosslinked
with BIS (A), or in a 9% polyacrylamide gel crosslinked with DATD (B). Ten differ-
ent polypeptides were identified in both slab gels and arbitrarily designated l-10.
Those migrating markedly different in the two slab gels are labeled with large arrows,
indicating the direction of deviation in B relative to A.
material in identical pH gradients, but on two different slab gels. Proteins of
interest are marked l-10. The relative positions of spots 1, 2, 6, 7, and 8 vary
considerably when N,Nâ€™-diallyltartaramide (DATD) is used as crosslinker
instead of NJâ€™-methylene-his-acrylamide (BIS). Such variations are also
observed for other proteins in other areas of the slab gel (results not shown).
The discrepancies some time ago about the M,s of certain HSV proteins can
thus be explained by the fact that different crosslinkers were used.
2.1.4. Detection of Separated Proteins
by Staining, Autoradiography or Fluorography
The sensitivity of silver staining is 50-100 times higher than that of
Coomassie blue (26), and comparable to that of autoradiography. Loading of
10-20 pg of total protein per tube gel is usually enough to visualize a large
number of spots of varying intensity. Adjustment of the incubation period for
optimal staining sometimes is a tine balance between the intensities ofthe spots
and background staining. Radioactive labeling of the proteins has several
Haarr and Langeland
advantages. First, the loading on the gel depends on the amount of radioactiv-
ity rather than of protein. Second, the same gel can be exposed to films for a
variety of periods such that heavily labeled and weakly labeled bands or spots
are distinct after short and long exposures, respectively. Third, in our hands the
background is clearer than after silver staining.
The low-energy p-emission from [3H] is almost undetectable unless special
films with high sensitivity are used. Alternattvely, the radtoactlve signal is
converted into scmtillation by mfusion of the gel with a scintillant. In the origr-
nal procedure of Bonner and Laskey (27) the hydrophobic compound 2-5
dtphenyloxazole (PPO) was dissolved in drmethylsulfoxide (DMSO) and
precipitated in the gel by addition of water. The dried gels are exposed to film
at -80Â°C. DMSO, however, is toxic and penetrates skm as well as mucous
membranes both m the liquid form and as vapor. Less toxic or atoxic scintillants
prepared for the same purpose are available commercially (Amplify from
Amersham [Arlington Heights, IL] and En3Hance from New England Nuclear
[Beverly, MA]). Signals from P-emitting isotopes with energies slightly higher
than [3H] ([ 14C] and [35S]) are increased by a factor of 10 by this technique.
To increasethe incorporation of [35S]-methtonineinto HSV proteins the concen-
tration of methionine in the medium is reduced to 20% (28). Labeling for periods
shorter than 2-3 h can be performed m PBS. He et al. (29) compared the mcorpora-
non of [35S]-methionineand [ I4 Cl-amino acids into proteins in mouse fibroblasts
durmg a period of 24 h, and found rather surprisingly that the former method was
three to four tunes better. A few spots,however (3-4% of the total), were labeled
with [14C]and not wtth [35S]Glycoprotein L is oneexample of a HSV protein that is
labeled wtth [35S]-cysteme rather than with [35S]-methionine(30).
2.1.5. Detection of Protein Processing
Many of the herpes simplex vu-us proteins are processed further either con-
comitant with the translation or afterward. Eleven glycoproteins (gB, gC, gD,
gE, gG, gH, g1, gJ, gK, gL, and gM) are described so far (30-36). Most, prob-
ably all, of the tegument proteins (at least 11) are phosphorylated (reviewed in
ref. 33). Kinase activities are induced by HSV- 1 in infected cells (3 7), and the
nuclear kmase encoded by gene ULl 3 is also present m virtons (3 7-39) The
structural proteins VP24 and VP2 1 and the scaffolding protein VP22a, are gen-
erated by proteolytic cleavage of precursor protems from genes UL26 and
UL26.5, respectively (4044) A tegument protein (encoded by gene U,ll) is
myristylated (45,46). Another tegument protein, VP22, seems to be modified
by poly(ADP-ribosyl)ation (471, and the nnmediate+arly regulatory proteins
are subJected to both adenylation and guanylation (48).
Glycosylation is usually detected by the incorporation of [14C] glucosamine
or [3H]-mannose added directly to the medium. A number of different com-
Analysis of HSV Polypeptides 103
pounds mhibrt specific steps durmg glycosylatron. Tumcamycm blocks N-linked
glycosylation (49,50), whereas monensin impairs Golgi apparatus function, i.e.,
both O-linked glycosylation and the processing of N-linked moieties (51). Car-
bohydrate groups already bound to the proteins are cleaved specifically with
different enzymes (52,53). High-mannose ohgosaccarides are sensitive to
endoglycosidase H (endo+-N-acetyl glucosaminidase, endo H) until passage
through the Golgi, then resistant. Endoglycosidase F (endoglycosidase F/N-
glycosidase, endo F) cleaves N-linked oligosaccartdes specifically, both from
high mannose and complex glycoprotems. O-linked carbohydrates are sensr-
tive to 0-glycanase (endo-a-N-acetylgalactosaminidase). Sialic acid, which is
charged, is removed by neuraminidase.
Phosphoproteins are labeled with radioactive inorganic phosphate ([32P-P,J).
We find that the mcorporation is markedly increased by dialysis of the serum
and reducing the phosphate concentration in the medium to 10%. Beta-emis-
sion from [32P]has a relatively high energy capable of causing damage of dif-
ferent molecules, including nucleic acids. Thus, m the presence of 50-500 pCi/
35-mm dish we observed that DNA synthesis was reduced by approx 15% at
6-7 h after infection and addition of isotope, and by 50% at 12 h
Myristylation takes place during translation as covalent bmding of myrrstic
acid to a consensus sequence m the N-termmus (54). It ISdetected by the incor-
poration of [3H]-myristic acid, and has been described for a large number of
different proteins in various viruses (55-62). The only HSV product known so
far to belong to this group is encoded by gene UL 11 (45,46). Protem precursors
are often labeled in pulses of 10-30 min with the radioactive compound, and the
final products detected after a subsequent chase. Some posttranslational modi-
fications, however, may occur even during the shortest possible pulse, Systems
for m vitro translation of mRNAs are normally so different from infected cell
cultures that posttranslational modrfication either is deficient or absent. In vitro
translated proteins thus may be considered to be more or less identical to the
primary translation products. Some large mRNAs may not be translated effi-
ciently m vitro. Removal of posttranslationally added monies like carbohy-
drate or phosphate groups usually will result in a mobility shift during
electrophoresis such that the precursor-product relationship is established.
Specific antibodies used m immunopreciprtation or in Western blot may
unambiguosly demonstrate such relationships.
Cleavage of HSV polypeptides owing to the unspecific action of cellular pro-
teases should be avoided by adding protease inhibitors, for example, 0.1 mM
phenylmethylsulfonyl fluoride (PMSF), and by keeping the preparations on ice.
Figures 4 and 5 show examples of processing of HSV-proteins detected by
2D gel electrophoresis in combination with pulse-chase, in vitro translation,
and enzymatic treatment.
Haarr and Lange/and
â€˜iN VITRO ThANSL.
Fig. 4. Posttranslational modifications of proteins detected by 2D gel electrophore-
sis. BHK cells were infected with HSV-1. The proteins were labeled with [35S]-
methionine either by in vitro translation of isolated RNA, by a 30-min pulse at 5 h post
infection, or by a 30-min pulse followed by chase for 12 h. The slab gels contained 9%
polyacrylamide crosslinked with DATD. In the right panels the glycoproteins were
labeled with [3H]-mannose using a 30-min pulse only, or a pulse followed by chase for
6 h. Gradients of 512.5% polyacrylamide crosslinked with BIS were used in the slab
gels. Processing is indicated for the 65 K DNA-binding protein (65KDnP), gB, gC, gD,
and for the polypeptides arbitrarily designated a, b, c, and d. VP 16 and actin (A) are
unmodified and used as markers.
Analysis of HSV Polypeptides
w i T !-? E M L v Y E WITH ENZYME
Fig. 5. Enzymatic removal of moities added by posttranslational modification. BHK
cells were infected with HSV-1 and the proteins labeled with either [â€˜*PI-Pi (left pan-
els or [35S]-methionine (right panels). The phosophoprotein preparation was incu-
bated for 30 min at 37Â°C with either 15 U/mL of phosphodiesterase (EC 220.127.116.11.) or
with water. The slab gels contained 9% polyacrylamide crosslinked with DATD. Other
HSV- 1-infected cells were labeled with [35S]-methionine and the proteins incubated
for 14 h at 37Â°C with either 0.02 U neuraminidase (EC 18.104.22.168.) in 40 pL total vol-
ume, or with water. Two-dimensional gel electrophoresis was performed using a
mixture of 20% ampholines pH 3.5-10.0 and 80% ampholines pH 5.0-7.0 such
that the region pH 5.0-7.0 was expanded. The slab gels contained 5-12.5% polyacry-
lamide crosslinked with BIS. Some virus-induced proteins are indicated as well as the
polypeptides arbitrarily designated 1-7.