Anti-influenza activity of phenethylphenylphthalimide analogs derived from thalidomide

Article abstract of DOI:10.1016/j.bmc.2010.05.035

Swine-origin influenza A virus has caused pandemics throughout the world and influenza A is regarded as a serious global health issue. Hence, novel drugs that will target these viruses are very desirable. Influenza A expresses an RNA polymerase essential for its transcription and replication which comprises PA, PB1, and PB2 subunits. We identified potential novel anti-influenza agents from a screen of 34 synthesized phenethylphenylphthalimide analogs derived from thalidomide (PPT analogs). For this screen we used a PA endonuclease inhibition assay, a PB2 pathogenicity-determinant domain-binding assay, and an anti-influenza A virus assay. Three PPT analogs, PPT-65, PPT-66, and PPT-67, were found to both inhibit PA endonuclease activity and retard the growth of influenza A, suggesting a correlation between their activities. PPT-28 was also found to inhibit the growth of influenza A. These four analogs have a 3,4-dihydroxyphenethyl group in common. We also discuss the possibility that 3,4-dihydroxyphenethyl group flexibility may play an important functional role in PA endonuclease inhibition. Another analog harboring a dimethoxyphenethyl group, PPT-62, showed PB2 pathogenicity-determinant domain-binding activity, but did not inhibit the growth of the virus. Our present results indicate the utility of the PA endonuclease assay in the screening of anti-influenza drugs and are therefore useful for future strategies to develop novel anti-influenza A drugs and for mapping the function of the influenza A RNA polymerase subunits.

Full text of DOI:10.1016/j.bmc.2010.05.035

Bioorganic & Medicinal Chemistry 18 (2010) 5379–5390
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Anti-influenza activity of phenethylphenylphthalimide analogs derived
from thalidomide
Yuma Iwai a, , Hitoshi Takahashi b, , Dai Hatakeyama a, Kazunori Motoshima c, Minoru Ishikawa c,
Kazuyuki Sugita c, Yuichi Hashimoto c, Yuichi Harada b, Shigeyuki Itamura b, Takato Odagiri b,
Masato Tashiro b, Yoshihisa Sei d, Kentaro Yamaguchi d, Takashi Kuzuhara a,
*
a Laboratory of Biochemistry, Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770-8514, Japan
b Center for Influenza Virus Research, National Institute of Infectious Diseases, Gakuen 4-7-1, Musashimurayama-shi, Tokyo 208-0011, Japan
c Institute of Molecular & Cellular Biosciences, The University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan
d Analytical Chemistry Laboratory, Faculty of Pharmaceutical Sciences at Kagawa Campus, Tokushima Bunri University, Sanuki-City, Kagawa 769-2193, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Swine-origin influenza A virus has caused pandemics throughout the world and influenza A is regarded as
a serious global health issue. Hence, novel drugs that will target these viruses are very desirable. Influ-
enza A expresses an RNA polymerase essential for its transcription and replication which comprises
PA, PB1, and PB2 subunits. We identified potential novel anti-influenza agents from a screen of 34 syn-
thesized phenethylphenylphthalimide analogs derived from thalidomide (PPT analogs). For this screen
we used a PA endonuclease inhibition assay, a PB2 pathogenicity-determinant domain-binding assay,
and an anti-influenza A virus assay. Three PPT analogs, PPT-65, PPT-66, and PPT-67, were found to both
inhibit PA endonuclease activity and retard the growth of influenza A, suggesting a correlation between
their activities. PPT-28 was also found to inhibit the growth of influenza A. These four analogs have a 3,4-
dihydroxyphenethyl group in common. We also discuss the possibility that 3,4-dihydroxyphenethyl
group flexibility may play an important functional role in PA endonuclease inhibition. Another analog
harboring a dimethoxyphenethyl group, PPT-62, showed PB2 pathogenicity-determinant domain-binding
activity, but did not inhibit the growth of the virus. Our present results indicate the utility of the PA endo-
nuclease assay in the screening of anti-influenza drugs and are therefore useful for future strategies to
develop novel anti-influenza A drugs and for mapping the function of the influenza A RNA polymerase
subunits.
Received 21 April 2010
Revised 12 May 2010
Accepted 13 May 2010
Available online 20 May 2010
Keywords:
Influenza
Thalidomide
RNA polymerase
Phenethylphenylphthalimide
Endonuclease
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
thalidomide is a multi-target drug and is thought to be useful as a
template in the development of other biologically active com-
Thalidomide, a hypnotic/sedative drug, was originally launched
in the 1950s but was subsequently withdrawn from the market in
the 1960s because of its teratogenic properties.1,2 Thalidomide has
been subsequently shown however to be useful in the treatment of
Hansen’s disease, multiple myeloma, cancer, rheumatoid arthritis,
graft-versus-host diseases, and acquired immunodeficiency syn-
drome.1–4 Pharmacologically, thalidomide has anti-cachexia, anti-
inflammatory, anti-tumor-promoting, anti-angiogenic, tumor cell
invasion-inhibiting, anti-viral, and hypoglycemic activities. Hence,
pounds. Indeed, Hashimoto and colleagues have previously devel-
oped various thalidomide analogs2,3,5–7 and based on the target
molecules have created tumor necrosis factor-production-regulat-
ing agents, cyclooxygenase inhibitors, nitric oxide synthase inhib-
itors, histone deacetylase inhibitors, anti-angiogenic agents, and
tubulin polymerization inhibitors.2,3,5–7 These researchers have
also developed the structure of thalidomide based on its pharma-
cological effects and thereby produced androgen antagonists, pro-
gesterone
antagonists,
cell
differentiation
inducers,
aminopeptidase inhibitors, thymidine phosphorylase inhibitors,
l
-calpain inhibitors, a-glucosidase inhibitors, and nuclear liver X
Abbreviations: PPT, phenethylphenylphthalimide; IPTG, isopropyl b-D-thioga-
receptor antagonists.2,3,5–7
lactopyranoside; ESI, electrospray ionization; MS, mass spectrometry; FT-ICRMS,
Fourier transform ion cyclotron resonance mass spectrometer; MDCK, Madin–
Darby canine kidney.
* Corresponding author. Tel.: +81 88 602 8477; fax: +81 88 655 3051.
E-mail address: kuzuhara@ph.bunri-u.ac.jp (T. Kuzuhara).
These authors contributed equally to this work.
In 1918, an influenza A pandemic caused ten million deaths
worldwide8 and strategies to prevent any future expansions of this
virus are therefore an important endeavor.9,10 The avian H5N1
influenza A virus is highly pathogenic to humans11 and the emer-
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.05.035

5380
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
gence of a new strain of this virus in 2009, the swine-origin A/
H1N1 pdm influenza virus (SOIV), emphasizes this issue further
as it has become a serious global health issue.12 Although inhibi-
tors of influenza A such as the neuraminidase-like compound osel-
tamivir are widely used as anti-viral drugs,13,14 some adverse
effects of these agents and also the emergence of viral strains that
are resistant to these drugs have now been reported.15–17
PPT analogs could inhibit this activity. In the assay, we incubated
0.1
10
l
l
M of recombinant PA endonuclease domain with both 1 and
M of each PPT analog (Fig. 1B). PPT-65 and PPT-66 significantly
M (Fig. 2) and
inhibited the digestion of M13mp18 at a dose of 10
l
this is the first evidence that agents based on the thalidomide skel-
eton can inhibit the influenza A virus endonuclease. PPT-67 also
showed weak inhibitory activity in this regard but none of the
other 31 PPT analogs tested has any effects in the assay (Fig. 2).
Interestingly, we found that PPT-65, PPT-66, and PPT-67 all contain
a 3,4-dihydroxyphenethyl group (Fig. 1B), indicating that this moi-
ety is important for the inhibition of PA endonuclease (Figs. 1B and
2). A detailed analysis of the structure–function relationships of
the PPT analogs is provided later in Section 3.
For the prevention and control of influenza outbreaks, the
development of novel anti-viral drugs that are not based on neur-
aminidase inhibition is now regarded as critical.12 The influenza A
genome consists of a segmented single stranded RNA (À) and its
transcription and replication require the activity of a highly con-
served RNA-dependent RNA polymerase.18,19 This polymerase is
essential for the influenza A virus to propagate and thus represents
a very promising target for anti-viral drug development. The influ-
enza A virus RNA-dependent RNA polymerase is composed of three
subunits, PA, PB1, and PB2, and synthesizes viral mRNAs using
short capped primers derived from host cellular pre-mRNAs
cleaved by the viral endonuclease.18,19 Yuan et al. and Dias et al.
have shown that the N-terminal domain of the PA subunit contains
the endonuclease active site and that this domain also harbors
RNA/DNA endonuclease activity.20–22 The PB2 subunit includes
K627, which plays a role in the high pathogenicity and host range
restriction of the virus.23–26 We and others have elucidated the ter-
tiary structure of the C-terminal domain (627 domain) of PB2 by X-
ray crystallography and identified a unique loop and basic groove
proximal to the K627 residue.27,28 D701 N has also been shown
to be associated with the viral pathogenicity levels and is also con-
tained in this domain. Hence, we speculated that the PA endonu-
clease and PB2 627 domains would be very effective targets in
the development of novel anti-influenza A drugs.
2.3. Binding of PPT analogs to the PB2 627 domain
As another in vitro selection, we next screened for PB2 627 do-
main (residues 535–759)-binding factors among PPT analogs. This
domain has been reported to be associated with viral pathogenicity
and host range restriction,25–28 and we speculated that it would
play an important role and that any chemical that bound to this
domain would have anti-viral effects. We established an assay sys-
tem to detect the PB2 627 domain protein by ESI-MS.30,31 In this
analysis, molecular weights are presented as a number divided
by the charge number, which in the case of the PB2 627 domain
is 25146. As shown in Figure 3A, peaks with sodium adducts ap-
peared with the mass to charge (m/z), which corresponds to the
molecular weight of the PB2 627 domain (Fig. 2A). We thus con-
cluded that these peaks are derived from an intact PB2 627 domain
protein. We next mixed and analyzed 1–3
lM of PB2 627 domain
protein with 50 M of each PPT analog to detect whether binding
l
Our preliminary results in this regard suggested that phen-
ethylphenylphthalimide (PPT) analogs derived from thalidomide
are possible lead compounds (Supplementary Fig. 1) and we thus
screened a further cohort of PPT analogs using a PA endonuclease
inhibition assay, PB2 627 domain-binding assay and anti-influenza
A virus assay. The results of these assays and also of our analysis of
the structure–activity relationships of PPT analogs are described.
occurred. When PB2 627 domain mixed with PPT-62 (Fig. 3F), addi-
tional peaks other than the PB2 627 domain were appeared. The
differences of m/z values between additional peaks and those de-
rived from PB2 627 domain were 43.0 and 42.7 (Fig. 3F). When
multiplied by nine, these values become 387 and 384 which are al-
most same as the molecular weight of PPT-62 (387). In case of the
other charge, especially e = 8, shifted peaks were also observed
(Supplementary Fig. 2). This indicates that PPT-62 directly associ-
ates with the PB2 627 domain. No binding was observed for several
other PPT analogs (Fig. 3B–G), indicating that the association of
PPT-62 and PB2 627 is specific. Two kinds of bound form of PB2
627 domain, (PB2 627 domain + PPT-62) and (PB2 627 domain + t-
wo PPT-62), were observed indicating that two molecules of PPT-
62 bind to this domain. This may be because two PPT-62 molecules
2. Results
2.1. Synthesis of PPT analogs
PPT derivatives were synthesized as previously described29 in
which diphenylethene derivatives were prepared as E/Z mixtures
via a Wittig reaction of nitrobenzaldehyde with appropriately
substituted benzyl ylide (Fig. 1A). After reduction of the nitro
group and/or olefin moiety of the adducts, PPT derivatives were ob-
tained by condensation with phthalic anhydride (Fig. 1A). We de-
noted our PPT analog series as PPT-n (n = sequential numbering
from 1; Fig. 1B). We obtained 37 PPT analogs, three of which
(PPT-108, -109, and -112) could not be dissolved in the aqueous as-
say buffer. We then adopted a strategy for developing a novel anti-
influenza A drug which involved two initial in vitro screenings for
the inhibition of PA endonuclease activity and a PB2 627 domain-
binding assay. After these in vitro selections, we performed an
anti-virus assay.
interact with each other via
contains an aromatic ring.
p–p bonding because this chemical
2.4. Inhibition of influenza A by PPT analogs
We performed an influenza A virus assay in a cell culture sys-
tem for the four PPT analogs that selected by these two in vitro as-
says (PPT-62, -65, -66, and -67), and also include one further PPT
analog (PPT-28) containing a 3,4-dihydroxyphenethyl group, and
additional analogs (PPT-68 and -125) for comparison. MDCK cells
undergo cytopathy following infection by influenza A in culture
and we used this system to examine whether our selected PPT ana-
logs could inhibit growth of this virus. Various concentrations
(0.63–80 lM) of each PPT analog and 100 TCID50 (50% of the infec-
2.2. Inhibition of PA endonuclease by PPT analogs
tious dose) of influenza A virus were premixed, and then added to
the culture medium of the MDCK cells. We performed this assay
three times and thereby determined the viability of the MDCK cells
(Fig. 4). DMSO alone did not inhibit the viral growth whereas PPT-
65 and -66 did so at a dose of around 40 lM (Fig. 4). PPT-67 also
exerted these effects, but was more toxic to the cells (Supplemen-
In the in vitro PA endonuclease assay, we expressed and puri-
fied a recombinant PA endonuclease domain (1–220 residues) of
the influenza A virus RNA polymerase as previously described.22
The PA endonuclease domain digests M13mp18 ssDNA in vitro22
(Fig. 2, lanes 1 and 2) and we examined which of our 34 soluble
tary Fig. 3). This is the first evidence that compounds with a

Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
5381
R2
R1
A
R2
R1
a
CHO
O2N
Ph3P
O2N
2
4''
3''
R
X
X
R2
R1
O
2' 3'
X
X
b or c
d
R1
N
H2N
4'
O
X = H, Cl
R1, R2 = H or OMe
2
4''
3''
R
X
X
O
2' 3'
X
X
e
R1
N
4'
O
X = H, Cl
R1, R2 = H or OH
4''
3''
R2
R1
B
4''
Cl
Cl
R2
R1
O
2' 3'
4'
O
N
O
Cl
Cl
2' 3'
N
3''
4'
O
single (s) or
position double bond (d)
single (s) or
R1
R2
R1
H
OMe
OH
R2
Compound
Compound
position double bond (d)
PPT2P
PPT62
PPT65
2'
2'
2'
s
s
s
H
OMe
OH
H
OMe
OH
PPT43
PPT59
PPT67
2'
2'
2'
s
s
s
H
OMe
OH
3'
3'
s
s
OMe
OH
OMe
OH
3'
3'
3'
s
s
s
H
H
PPT63
PPT66
PPT61
PPT60
PPT68
OMe
OH
OMe
OH
4'
4'
s
s
H
OMe
H
OMe
PPT21
PPT32
4'
4'
4'
s
s
s
H
OMe
OH
H
OMe
OH
PPT22
PPT27
PPT28
4'
4'
d
d
H
OH
H
OH
PPT108
PPT113
3'
3'
3'
d
d
d
H
OMe
OH
H
OMe
OH
PPT84
PPT95
PPT97
4'
4'
4'
4'
s
s
s
s
OMe
H
OH
H
H
PPT121
PPT123
PPT125
PPT137
OMe
H
OH
4'
4'
d
d
H
OMe
H
OMe
PPT109
PPT112
3'
3'
3'
3'
s
s
s
s
OMe
H
H
PPT91
PPT80
PPT94
PPT85
OMe
H
Cl
Cl
OH
H
O
Z
3'
Compound position
Cl
Cl
OH
N
4'
PPT86
PPT87
3'
4'
4'
4'
4'
4'
s
s
s
s
OMe
H
OH
H
H
PPT122
PPT124
PPT136
PPT138
OMe
H
O
OH
Figure 1. (A) Scheme of the chemical synthesis of PPT analogs. Reagents and conditions: (a) K2CO3, 18-crown-6, CH2Cl2, reflux; (b) H2, 10%, Pd/C, EtOAc, rt; (c) SnCl2Á2H2O,
EtOAc, reflux; (d) phthalic anhydride or tetrachlorophthalic anhydride, neat, 200 °C; (e) BBr3, CH2Cl2, 0 °C. (B) Chemical structures of the synthesized PPT analogs screened for
anti-influenza A activities.
phthalimide skeleton derived from thalidomide possess anti-influ-
enza A virus activity. PPT-62, -68, and -125 did not inhibit the viral
growth (Fig. 4) nor inhibit PA endonuclease activity (Fig. 2). Hence,
a correlation exists between the inhibition of PA endonuclease and
inhibition of influenza A. Interestingly, we also found that PPT-28,
which contains a 3,4-dihydroxyphenethyl group, inhibits the influ-
enza virus but not via the targeting of the PA endonuclease. Table 1
lists the EC50 and CC50 values for these analogs which were calcu-
lated from the data shown in Figure 4. The average EC50 for PPT-28,
-65, and -66 was 24, 48, and 26
lM, and the CC50 of PPT-28 and -65
was above 80 M, respectively. These compounds are therefore
l
putative candidate anti-influenza A drugs themselves and may also
have utility as seeds for the development of such agents.
3. Discussion
We demonstrate that three PPT analogs, PPT-65, PPT-66, and
PPT-67, inhibit PA endonuclease and the viral growth. Significantly,

5382
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
PA
endonuclease
domain
-
+
61
10
62
10
63
10
65
10
66
10
67
10
80
10
chemicals PPT-
2P
10
21
10
22
10
27
10
28
10
32
10
43
10
59
10
60
10
0
0
1
1
1
8
1
1
1
1
1
1
1
1
1
1
1
1
1
conc. (µM)
substrate
degradated
lane
1
2
1
3
4
1
5
6
1
7
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
PA
endonuclease
domain
+
113
10
84
10
85
10
86
10
87
10
91
10
94
10
95
10
97
10
121
10
122
10
123
10
124
10
125
10
136
10
137
10
138
10
chemicals PPT-
conc. (µM)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
substrate
degradated
lane
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68
Figure 2. Screening of PPT analogs for anti-influenza A activity using a PA endonuclease assay. Effects of various PPT analogs upon the endonuclease activity of the N-terminal
domain of the PA subunit of influenza A virus RNA-dependent RNA polymerase. The recombinant N-terminal domain of PA was added to each reaction at 0.35 g/100 l. A
l
l
zero control (no PA domain added) was also assayed. PPT analogs were added at 1 or 10 lM and M13mp18 was used as the substrate.
all of these analogs contain a 3,4-dihydroxyphenethyl group at the
ortho or meta position of the N-phenyl moiety (Fig. 1B). The PPT
analogs possessing methoxyphenethyl group(s) (PPT-27, -32, -59,
-60, -62, -63, -80, -91, -95, -121, -122, -123, and -124), a mono-
hydroxyphenethyl group (PPT-85, -94, -125, -136, -137, and -
138) or an unsubstituted phenethyl group (PPT-21, -22, -43, -61,
-84, -86, -87 and PPT2P) (Fig. 1B) did not inhibit PA endonuclease
activity (Fig. 2). This suggests a central functional role of the 3,4-
dihydroxyphenethyl moiety for this inhibition. Previously, we have
reported that catechins32 can inhibit PA endonuclease22 and it has
been shown also that these molecules have anti-influenza activ-
ity.33 Significantly, the catechins also contain a 3,4-dihydroxy-
phenyl group (Supplementary Fig. 4), confirming that this moiety
plays a role in the inhibition of PA endonuclease and in suppressing
the growth of influenza (Fig. 5).
PPT-28 possesses a 3,4-dihydroxyphenethyl group at the para
position (Fig. 1B) and does not inhibit PA endonuclease (Fig. 2),
indicating that introduction of 3,4-dihydroxyphenethyl group into
the ‘ortho or meta’ position of the N-phenyl moiety be critical for
this activity. Moreover, conversion of the 3,4-dihydroxyphenethyl
group to a 3,4-dihydroxycinnamyl group, that is, as seen for PPT-
97 and -113 (Fig. 1B), results in the loss of PA endonuclease-inhib-
itory activity (Fig. 2), also indicating the importance of the flexible
3,4-dihydroxyphenethyl group for this process. A freely rotating
3,4-dihydroxyphenyl group might also play an important func-
tional role in this process. In combination with the importance of
the position (ortho or meta of the N-phenyl moiety) at which the
3,4-dihydroxyphenethyl group is introduced, the spatial position
of the two hydroxyl groups of the 3,4-dihydroxyphenethyl moiety
in the molecule seem also to be critical for this activity. PPT-68 has
a 3,4-dihydroxyphenethyl group but shows little or no inhibition of
PA endonuclease (data not shown). We suggest therefore that the
four chloride groups introduced into the phthalimide moiety inhi-
bit or weaken this activity as PPT-67 also possesses a tetrachlor-
ophthalimide group and shows weaker inhibition than PPT-65 or
PPT-66 (Figs. 1 and 2). Although we have not yet investigated other
functional groups derived from catechin, that is, 3,4,5-trihydroxy-
phenethyl and/or 3,5-dihydroxyphenethyl groups among others,
our current results indicate that the development of chemicals
with a 3,4-dihydroxyphenethyl group will likely produce more
effective PA endonuclease inhibitors and better anti-influenza A
drugs.
PPT-62 binds to the PB2 pathogenicity-determining 627 domain
and harbors a dimethoxyphenethyl group (Figs. 1B and 3). In con-
trast, PPT-21 and PPT-43 do not possess a dimethoxyphenethyl
group (Fig. 1B) and did not show binding to this domain (Fig. 3).
PPT-62 is very similar to PPT-67 except for a methoxy group
(Fig. 1B) but as PPT-67 binds only weakly to the PB2 627 domain
(Fig. 3), the dimethoxy group of PPT-62 is likely to be important
for this domain-binding. PPT-59 is identical to PPT-62 other than
four additional chlorides (Fig. 1B) and also binds only weakly to
the PB2 627 domain (Fig. 3), suggesting that these chlorides inhibit
this binding. Based on these results, the development of com-
pounds possessing a dimethoxyphenethyl group will likely lead
to the production of drugs with a higher affinity for the PB2 627
domain.
Our current data show that PPT analogs can be used as tools
to investigate the functional roles of influenza A virus RNA poly-
merase domains. Based on our results (summarized in Fig. 6),
the PB2 pathogenicity-determinant domain may have a regula-
tory function rather than an essential one. Also, the correlation
between the inhibition of PA endonuclease and suppression of
the influenza A virus confirmed the importance of this enzyme
for viral growth. Our results also indicate that the PA endonucle-
ase assay is a useful screening tool for novel anti-influenza A
drugs as the information obtained correlates well with virus

Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
5383
Figure 3. (A) Establishment of a detection system for factors that bind to the PB2 627 domain protein using FT-ICRMS. Vertical axis shows the intensity of the protein peaks.
The horizontal axis indicates the m/z values. (B–G) Binding of PPT analogs to the PB2 627 domain. Direct binding analysis of PPT analogs to the PB2 627 domain using FT-
ICRMS. The PB2 627 domain and PPT analogs were mixed and applied to FT-ICRMS. The vertical axis shows the intensity of the peaks. The horizontal axis indicates the m/z
values. (B) PB2 627 domain only. The PB2 627 domain was also incubated with (C) PPT-21; (D) PPT-43; (E) PPT-59; (F) PPT-62 and (G) PPT-67.

5384
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
Figure 4. Inhibition of influenza A viral growth by PPT analogs. MDCK cells were treated with (red bar) or without (blue bar) influenza A virus. The cells were then treated
with various concentrations of PPT analogs (0.63–80 M), some of which suppressed viral-induced cell death. DMSO was used as the PPT solvent. The vertical and horizontal
l
axes indicate cell viability (%) and the concentration of PPT analogs, respectively. Experiments were performed in triplicate, and the average values with standard deviations
are indicated.
growth and this method is easier and thus more convenient to
perform than other screening approaches using the virus itself.
The effective doses of PPT analogs are high at present, but as
the tertiary structures of the PA endonuclease and PB2 627 do-
mains have already been determined,20,21,27,28 docking simula-
tions in silico34 can be used for the further development and
optimization of these compounds. We conclude finally that the
chemical and biochemical information presented herein will be

Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
5385
Table 1
EC50 and CC50 values for the anti-influenza A activities of the indicated PPT analogs
Virus (lM)
DMSO
PPT-28
PPT-62
PPT-65
PPT-66
PPT-68
PPT-125
(+)
EC50
(À)
(+)
EC50
(À)
(+)
EC50
(À)
(+)
EC50
(À)
(+)
EC50
(À)
(+)
EC50
(À)
(+)
EC50
(À)
CC50
CC50
CC50
CC50
CC50
CC50
CC50
1st
2nd
3rd
>80
>80
>80
>80
>80
>80
30
25
16
>80
>80
>80
>80
>80
>80
>80
>80
>80
60
37
>80
>80
>80
28
11
40
>80
48
57
>80
>80
>80
62
34
37
>80
>80
>80
>80
>80
>80
>80a
Average
>80
>80
24
>80
>80
>80
48b
>80
26
nd
>80
44
>80
>80
The EC50 and CC50
(lM) values were estimated from a dose-dependent curve in Figure 4.
nd, not determined.
a
The EC40 was around 50 (lM). The EC50 was not estimated due to the low viability of the cells exposed to this higher dose. The shapes of the curves calculated for the three
experiments are similar as shown in Supplementary Figure 5.
b
Average was calculated from the first and second experiments in this case. See Supplementary Figure 5.
4.1.2. N-(4-(2-Phenylethyl)phenyl)phthalimide: PPT-21
Mp 216.5–218.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.50 (dd, 2H, J = 5.5, 3.1 Hz), 7.36–7.29 (m, 6H),
7.23–7.19 (m, 3H), 3.01–2.94 (m, 4H). FAB-MS m/z: 327 [M]+,
328 [M+H]+. Anal. Calcd for C22H17NO2: C, 80.71; H, 5.23; N, 4.28.
Found: C, 80.71; H, 5.53; N, 4.27.
4.1.3. 4,5,6,7-Tetrachloro-N-[4-(2-phenylethyl)phenyl]phthali
mide: PPT-22
Mp 222.0–223.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.33–7.28 (m,
6H), 7.22–7.19 (m, 3H), 3.00–2.98 (m, 2H), 2.97–2.94 (m, 2H). FAB-
MS m/z: 463 [M]+, 464 [M+H]+, 465 [M+2]+, 466 [M+3]+, 467
[M+4]+, 468 [M+5]+. Anal. Calcd for C22H13Cl4NO2: C, 56.81; H,
2.82; N, 3.01. Found: C, 56.85; H, 3.02; N, 2.94.
4.1.4. 4,5,6,7-Tetrachloro-N-{4-[2-(3,4-dimethoxyphenyl)
ethyl]phenyl}phthalimide: PPT-27
Mp 220.0–221.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.28 (s, 4H),
6.78 (d, 1H, J = 8.5 Hz), 6.71 (dd, 1H, J = 8.5, 1.8 Hz), 6.62 (d, 1H,
J = 1.8 Hz), 3.85 (s, 3H), 3.82 (s, 3H), 2.96–2.93 (m, 2H), 2.90–2.87
(m, 2H). FAB-MS m/z: 523 [M]+, 524 [M+H]+, 525 [M+2]+, 526
[M+3]+, 527 [M+4]+, 528 [M+5]+. HRMS (FAB) calcd for
C
24H17Cl4NO4 522.9912; found: 522.9929 (M)+.
4.1.5. 4,5,6,7-Tetrachloro-N-{4-[2-(3,4-
dihydroxyphenyl)ethyl]phenyl}phthalimide: PPT-28
Mp 260.0–263.0 °C. 1H NMR (500 MHz, DMSO-d6) d: 8.70 (s,
1H), 8.62 (s, 1H), 7.38 (d, 2H, J = 8.5 Hz), 7.30 (d, 2H, J = 8.5 Hz),
6.64 (d, 1H, J = 2.1 Hz), 6.62 (d, 1H, J = 7.9 Hz), 6.47 (dd, 1H,
J = 7.9, 2.1 Hz), 2.88–2.85 (m, 2H), 2.75–2.72 (m, 2H). FAB-MS m/
z: 495 [M]+, 496 [M+H]+, 497 [M+2]+, 498 [M+3]+, 499 [M+4]+,
500 [M+5]+. Anal. Calcd for C22H13Cl4NO4Á1/3H2O: C, 52.52; H,
2.74; N, 2.78. Found: C, 52.47; H, 2.85; N, 2.51.
Figure 5. Consensus structure for the PPT analogs possessing anti-influenza A
activity. The 3,4-dihydroxyphenyl group is characteristic of catechin molecules
(shown as red) which also inhibit the influenza virus.
very useful for the future development of novel drugs against
the influenza A virus.
4.1.6. N-{4-[2-(3,4-Dimethoxyphenyl)ethyl]phenyl}
phthalimide: PPT-32
Mp 180.0–181.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.94 (dd, 2H,
J = 5.5, 3.1 Hz), 7.77 (dd, 2H, J = 5.5, 3.1 Hz), 7.33–7.28 (m, 4H),
6.79 (d, 1H, J = 7.9 Hz), 6.73 (dd, 1H, J = 7.9, 1.8 Hz), 6.63 (d, 1H,
J = 1.8 Hz), 3.85 (s, 3H), 3.82 (s, 3H), 2.95–2.92 (m, 2H), 2.90–2.87
(m, 2H). FAB-MS m/z: 387 [M]+, 388 [M+H]+. Anal. Calcd for
4. Experimental
4.1. Synthesis of PPT analogs
Synthesis of the PPT analogs is described in the text and de-
picted also in Figure 1A. Their chemical structures are depicted
in Figure 1B.
C24H21NO4Á1/5H2O: C, 73.72; H, 5.52; N, 3.58. Found: C, 73.98; H,
5.44; N, 3.64.
4.1.7. 4,5,6,7-Tetrachloro-N-[2-(2-phenylethyl)phenyl]phthal
imide: PPT-43
4.1.1. N-(2-(2-Phenylethyl)phenyl)phthalimide: PPT2P
Mp 90.8–91.2 °C. 1H NMR (500 MHz, CDCl3) d: 7.94–7.98 (m,
2H), 7.78–7.82 (m, 2H), 7.40 (dt, 1H, J = 7.5, 1.5 Hz), 7.33–7.37 (m
2H), 7.16–7.22 (m, 3H), 7.10–7.15 (m 1H), 7.04–7.08 (m, 2H),
2.78–2.89 (m, 4H). Anal. Calcd for C22H17NO2: C, 80.71; H, 5.23;
N, 4.28. Found: C, 80.85; H, 5.36; N, 4.20.
Mp 148.0–149.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.42–7.39 (m,
1H), 7.35–7.32 (m, 2H), 7.19–7.16 (m, 2H), 7.14–7.09 (m, 2H), 7.06
(d, 2H, J = 6.7 Hz), 2.87–2.84 (m, 2H), 2.78–2.75 (m, 2H). FAB-MS
m/z: 463 [M]+, 464 [M+H]+, 465 [M+2]+, 466 [M+3]+, 467 [M+4]+,

5386
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
Figure 6. The functional domains of influenza A virus RNA polymerase mapped with respect to their interaction with the PPT analogs. PPT-65, -66, and -67 inhibit PA
endonuclease activity and have anti-influenza A virus activity. PPT-62 binds to the PB2 pathogenicity-determinant 627 domain. PPT-28 has anti-influenza A virus activity.
Structure–function relationship analysis indicates that the 3,4-dihydroxyphenethyl group is important for the anti-influenza A activity of these analogs.
468 [M+5]+. Anal. Calcd for C22H13Cl4NO2: C, 56.81; H, 2.82; N, 3.01.
Found: C, 56.67; H, 2.92; N, 2.94.
(m, 2H), 2.91–2.88 (m, 2H). FAB-MS m/z: 523 [M]+, 524 [M+H]+,
525 [M+2]+, 526 [M+3]+, 527 [M+4]+, 528 [M+5]+. Anal. Calcd for
C
24H17Cl4NO4Á1/4H2O: C, 54.42; H, 3.33; N, 2.64. Found: C, 54.49;
4.1.8. 4,5,6,7-Tetrachloro-N-{2-[2-(3, 4-dimethoxyphenyl)ethyl]
phenyl}phthalimide: PPT-59
H, 3.36; N, 2.61.
Mp 157.0–158.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.41 (td, 1H,
J = 7.9, 1.8 Hz), 7.35 (td, 1H, J = 7.9, 1.8 Hz), 7.32 (dd, 1H, J = 7.3,
1.8 Hz), 7.15 (dd, 1H, J = 7.9, 1.8 Hz), 6.68 (d, 1H, J = 7.9 Hz), 6.61
(dd, 1H, J = 7.9, 1.8 Hz), 6.53 (d, 1H, J = 1.8 Hz), 3.81 (s, 3H), 3.75
(s, 3H), 2.80 (s, 4H). FAB-MS m/z: 523 [M]+, 524 [M+H]+, 525
[M+2]+, 526 [M+3]+, 527 [M+4]+, 528 [M+5]+. Anal. Calcd for
4.1.10. 4,5,6,7-Tetrachloro-N-[3-(2-phenylethyl)phenyl]phthal-
imide: PPT-61
Mp 187.0–187.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.42 (t, 1H,
J = 7.9 Hz), 7.29 (t, 2H, J = 7.9 Hz), 7.26–7.25 (m, 1H), 7.24 (dd,
2H, J = 7.9, 1.8 Hz), 7.20 (dd, 1H, J = 7.9, 1.2 Hz), 7.19 (d, 2H,
J = 7.9 Hz), 3.00–2.97 (m, 2H), 2.96–2.94 (m, 2H). FAB-MS m/z:
463 [M]+, 464 [M+H]+, 465 [M+2]+, 466 [M+3]+, 467 [M+4]+, 468
[M+5]+. Anal. Calcd for C22H13Cl4NO2: C, 56.81; H, 2.82; N, 3.01.
Found: C, 56.66; H, 2.91; N, 2.95.
C24H17Cl4NO4: C, 54.88; H, 3.26; N, 2.67. Found: C, 54.63; H,
3.37; N, 2.54.
4.1.9. 4,5,6,7-Tetrachloro-N-{3-[2-(3,4-dimethoxyphenyl)ethyl]
phenyl}phthalimide: PPT-60
4.1.11. N-{2-[2-(3,4-Dimethoxyphenyl)ethyl]phenyl}phthali-
mide: PPT-62
Mp 166.0–167.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.43–7.39 (m,
1H), 7.24–7.20 (m, 3H), 6.79 (d, 1H, J = 7.9 Hz), 6.71 (dd, 1H, J = 7.9,
1.8 Hz), 6.64 (d, 1H, J = 1.8 Hz), 3.86 (s, 3H), 3.83 (s, 3H), 2.97–2.94
Mp 103.0–105.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.95 (dd, 2H,
J = 5.5, 3.1 Hz), 7.80 (dd, 2H, J = 5.5, 3.1 Hz), 7.41–7.37 (m, 1H),

Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
5387
7.37–7.33 (m, 1H), 7.31 (dd, 1H, J = 7.3, 1.8 Hz), 7.20 (dd, 1H, J = 7.3,
4.1.18. 4,5,6,7-Tetrachloro-N-{3-[(1E)-2-phenylethenyl]
1.8 Hz), 6.69 (d, 1H, J = 7.9 Hz), 6.59 (dd, 1H, J = 7.9, 1.8 Hz), 6.50 (d,
1H, J = 1.8 Hz), 3.80 (s, 3H), 3.68 (s, 3H), 2.79 (s, 4H). FAB-MS m/z:
387 [M]+, 388 [M+H]+. Anal. Calcd for C24H21NO4: C, 74.40; H, 5.46;
N, 3.62. Found: C, 74.34; H, 5.59; N, 3.55.
phenyl}phthalimide: PPT-84
Mp 259.5–261.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.58–7.56 (m,
2H), 7.52–7.49 (m, 3H), 7.37 (t, 2H, J = 7.9 Hz), 7.31–7.28 (m, 2H),
7.15 (d, 1H, J = 16.8 Hz), 7.12 (d, 1H, J = 16.8 Hz). FAB-MS m/z:
461 [M]+, 462 [M+H]+, 463 [M+2]+, 464 [M+3]+, 465 [M+4]+, 466
[M+5]+. Anal. Calcd for C22H11Cl4NO2Á1/3H2O: C, 56.32; H, 2.51;
N, 2.99. Found: C, 56.55; H, 2.56; N, 2.98.
4.1.12. N-{3-[2-(3,4-Dimethoxyphenyl)ethyl]phenyl}
phthalimide: PPT-63
Mp 132.5–133.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.80 (dd, 2H, J = 5.5, 3.1 Hz), 7.43–7.39 (m, 1H),
7.28–7.26 (m, 2H), 7.19 (d, 1H, J = 7.3 Hz), 6.80 (d, 1H, J = 7.9 Hz),
6.73 (dd, 1H, J = 7.9, 1.8 Hz), 6.65 (d, 1H, J = 1.8 Hz), 3.85 (s, 3H),
3.83 (s, 3H), 2.99–2.95 (m, 2H), 2.93–2.89 (m, 2H). FAB-MS m/z:
387 [M]+, 388 [M+H]+. Anal. Calcd for C24H21NO4: C, 74.40; H,
5.46; N, 3.62. Found: C, 74.33; H, 5.49; N, 3.54.
4.1.19. 4,5,6,7-Tetrachloro-N-{3-[2-(4-hydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-85
Mp 210.0–212.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.41 (t, 1H,
J = 7.3 Hz), 7.24–7.20 (m, 3H), 7.03 (d, 2H, J = 8.5 Hz), 6.75 (d, 2H,
J = 8.5 Hz), 4.60 (br s, 1H), 2.95–2.92 (m, 2H), 2.89–2.86 (m, 2H).
FAB-MS m/z: 479 [M]+, 480 [M+H]+, 481 [M+2]+, 482 [M+3]+, 483
[M+4]+, 484 [M+5]+. Anal. Calcd for C22H13Cl4NO3: C, 54.92; H,
2.72; N, 2.91. Found: C, 54.68; H, 2.82; N, 2.83.
4.1.13. N-{2-[2-(3,4-Dihydroxyphenyl)ethyl]phenyl}phthal
imide: PPT-65
Mp 63.0–65.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.95 (dd, 2H,
J = 5.5, 2.4 Hz), 7.81 (dd, 2H, J = 5.5, 2.4 Hz), 7.41–7.38 (m, 1H),
7.36–7.31 (m, 2H), 7.18 (dd, 1H, J = 7.9, 1.2 Hz), 6.66 (d, 1H,
J = 7.9 Hz), 6.52 (d, 1H, J = 1.8 Hz), 6.49 (dd, 1H, J = 7.9, 1.8 Hz),
5.16 (s, 1H), 5.00 (s, 1H), 2.76 (m, 4H). FAB-MS m/z: 359 [M]+,
360 [M+H]+. HRMS (FAB) calcd for C22H17NO4 359.1158; found:
359.1127 (M)+.
4.1.20. 4,5,6,7-Tetrachloro-N-{3-[(1Z)-2-phenylethenyl]
phenyl}phthalimide: PPT-86
White powder from CH2Cl2/n-hexane. Mp 164.0–165.5 °C. 1H
NMR (500 MHz, CDCl3) d: 7.37–7.26 (m, 5H), 7.25–7.18 (m, 4H),
6.67 (d, 1H, J = 12.2 Hz), 6.59 (d, 1H, J = 12.2 Hz). FAB-MS m/z:
461 [M]+, 462 [M+H]+, 463 [M+2]+, 464 [M+3]+, 465 [M+4]+, 466
[M+5]+. Anal. Calcd for C22H11Cl4NO2: C, 57.05; H, 2.39; N, 3.02.
Found: C, 56.81; H, 2.53; N, 2.93.
4.1.14. N-{3-[2-(3,4-Dihydroxyphenyl)ethyl]phenyl}phthal
imide: PPT-66
4.1.21. 4,5,6,7-Tetrachloro-N-{4-[(1Z)-2-phenylethenyl]
phenyl}phthalimide: PPT-87
Mp 171.5–172.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 2.4 Hz), 7.81 (dd, 2H, J = 5.5, 2.4 Hz), 7.42 (t, 1H,
J = 7.9 Hz), 7.23 (dd, 1H, J = 7.9, 2.4 Hz), 7.22 (dd, 1H, J = 7.9,
1.8 Hz), 7.13 (m, 1H), 6.78 (d, 1H, J = 7.9 Hz), 6.63 (dd, 1H, J = 7.9,
1.8 Hz), 6.56 (d, 1H, J = 1.8 Hz), 5.66 (s, 1H), 5.16 (s, 1H), 2.94–
2.90 (m, 2H), 2.86–2.83 (m, 2H). FAB-MS m/z: 359 [M]+, 360
[M+H]+. Anal. Calcd for C22H17NO4: C, 73.53; H, 4.77; N, 3.90.
Found: C, 73.24; H, 4.84; N, 3.83.
White powder from CH2Cl2/n-hexane. Mp 155.5–157.5 °C. 1H
NMR (500 MHz, CDCl3) d: 7.58–7.49 (m, 1H), 7.38–7.18 (m, 8H),
6.67 (d, 1H, J = 12.8 Hz), 6.59 (d, 1H, J = 12.8 Hz). FAB-MS m/z:
461 [M]+, 462 [M+H]+, 463 [M+2]+, 464 [M+3]+, 465 [M+4]+, 466
[M+5]+. Anal. Calcd for C22H11Cl4NO2Á1/3H2O: C, 56.32; H, 2.51;
N, 2.99. Found: C, 56.35; H, 2.60; N, 2.96.
4.1.22. 4,5,6,7-Tetrachloro-N-{3-[2-(3-methoxyphenyl)
ethyl]phenyl}phthalimide: PPT-91
4.1.15. 4,5,6,7-Tetrachloro-N-{2-[2-(3,4-dihydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-67
Mp 150.0–151.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.42 (t, 1H,
J = 7.9 Hz), 7.26–7.25 (m, 2H), 7.23 (m, 1H), 7.20 (t, 1H,
J = 7.9 Hz), 6.78 (d, 1H, J = 7.3 Hz), 6.75 (dd, 1H, J = 7.9, 1.8 Hz),
6.72–6.72 (m, 1H), 3.78 (s, 3H), 3.00–2.96 (m, 2H), 2.94–2.91 (m,
2H). FAB-MS m/z: 493 [M]+, 494 [M+H]+, 495 [M+2]+, 496 [M+3]+,
497 [M+4]+, 498 [M+5]+. Anal. Calcd for C23H15Cl4NO3Á1/4H2O: C,
55.28; H, 3.13; N, 2.80. Found: C, 55.43; H, 3.11; N, 2.82.
Mp 227.0–229.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.41 (m, 1H),
7.36–7.32 (m, 2H), 7.14–7.13 (m, 1H), 6.66 (d, 1H, J = 7.9 Hz),
6.53 (d, 1H, J = 2.4 Hz), 6.49 (dd, 1H, J = 7.9, 1.8 Hz), 5.09 (s, 1H),
4.94 (s, 1H), 2.76 (s, 4H). FAB-MS m/z: 495 [M]+, 496 [M+H]+, 497
[M+2]+, 498 [M+3]+, 499 [M+4]+, 500 [M+5]+. Anal. Calcd for
C22H13Cl4NO4: C, 53.15; H, 2.64; N, 2.82. Found: C, 53.03; H,
2.88; N, 2.62.
4.1.23. 4,5,6,7-Tetrachloro-N-{3-[2-(3-hydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-94
4.1.16. 4,5,6,7-Tetrachloro-N-{3-[2-(3,4-dihydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-68
Mp 213.0–215.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.42 (t, 1H,
J = 7.9 Hz), 7.25–7.20 (m, 3H), 7.15 (t, 1H, J = 7.9 Hz), 6.76 (d, 1H,
J = 7.9 Hz), 6.68 (dd, 1H, J = 7.9, 2.4 Hz), 6.62 (t, 1H, J = 1.8 Hz),
2.98–2.95 (m, 2H), 2.92–2.88 (m, 2H). FAB-MS m/z: 479 [M]+,
480 [M+H]+, 481 [M+2]+, 482 [M+3]+, 483 [M+4]+, 484 [M+5]+. Anal.
Calcd for C22H13Cl4NO3: C, 54.92; H, 2.72; N, 2.91. Found: C, 54.68;
H, 2.99; N, 2.77.
Mp 208.0–210.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.43 (t, 1H,
J = 7.9 Hz), 7.24 (d, 1H, J = 7.3 Hz), 7.20 (m, 1H), 7.10 (s, 1H), 6.78
(d, 1H, J = 7.9 Hz), 6.63 (dd, 1H, J = 7.9, 1.8 Hz), 6.58 (d, 1H,
J = 1.8 Hz), 5.53 (s, 1H), 5.15 (s, 1H), 2.93–2.91 (m, 2H), 2.85–2.82
(m, 2H). FAB-MS m/z: 495 [M]+, 496 [M+H]+, 497 [M+2]+, 498
[M+3]+, 499 [M+4]+, 500 [M+5]+. Anal. Calcd for C22H13Cl4NO4Á2/
3H2O: C, 51.90; H, 2.84; N, 2.75. Found: C, 51.70; H, 3.03; N, 2.60.
4.1.24. 4,5,6,7-Tetrachloro-N-{3-[(1E)-2-(3,4-dimethoxyphenyl)
ethenyl]phenyl}phthalimide: PPT-95
4.1.17. 4,5,6,7-Tetrachloro-N-{3-[2-(4-methoxyphenyl)
ethyl]phenyl}phthalimide: PPT-80
Mp 248.0–249.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.56–7.54 (m,
2H), 7.49 (t, 1H, J = 7.9 Hz), 7.29–7.27 (m, 1H), 7.09 (d, 1H,
J = 16.2 Hz), 7.07–7.05 (m, 2H), 6.99 (d, 1H, J = 16.2 Hz), 6.87 (d,
1H, J = 7.9 Hz), 3.95 (s, 3H), 3.91 (s, 3H). FAB-MS m/z: 521 [M]+,
522 [M+H]+, 523 [M+2]+, 524 [M+3]+, 525 [M+4]+, 526 [M+5]+. Anal.
Calcd for C24H15Cl4NO4Á1/2H2O: C, 54.16; H, 3.03; N, 2.63. Found: C,
53.98; H, 3.03; N, 2.49.
Mp 169.0–170.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.41 (t, 1H,
J = 8.5 Hz), 7.23–7.22 (m, 2H), 7.09 (d, 2H, J = 8.5 Hz), 6.83 (d, 2H,
J = 8.5 Hz), 3.79 (s, 3H), 2.96–2.93 (m, 2H), 2.91–2.87 (m, 2H).
FAB-MS m/z: 494 [M+H]+, 495 [M+2]+, 496 [M+3]+, 497 [M+4]+,
498 [M+5]+. Anal. Calcd for C23H15Cl4NO3Á1/4H2O: C, 55.28; H,
3.13; N, 2.80. Found: C, 55.28; H, 3.09; N, 2.78.

5388
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
4.1.25. 4,5,6,7-Tetrachloro-N-{3-[(1E)-2-(3,4-dihydroxyphenyl)
ethenyl]phenyl}phthalimide: PPT-97
496 [M+3]+, 497 [M+4]+, 498 [M+5]+. HRMS (FAB) calcd for
23H15Cl4NO3 492.9806; found: 492.9834 (M)+.
C
Mp 257.0–260.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.51 (dt, 1H,
J = 7.9, 1.2 Hz), 7.50 (m, 1H), 7.47 (t, 1H, J = 7.9 Hz), 7.26–7.25 (m,
1H), 7.05 (d, 1H, J = 1.8 Hz), 7.00 (d, 1H, J = 16.5 Hz), 6.95 (dd, 1H,
J = 8.5, 1.8 Hz), 6.91 (d, 1H, J = 16.5 Hz), 6.85 (d, 1H, J = 8.5 Hz),
5.59 (br s, 1H), 5.54 (br s, 1H). FAB-MS m/z: 493 [M]+, 494
[M+H]+, 495 [M+2]+, 496 [M+3]+, 497 [M+4]+, 498 [M+5]+. HRMS
(FAB) calcd for C22H11Cl4NO4 492.9442; found: 492.9467 (M)+.
4.1.32. N-{4-[2-(4-Methoxyphenyl)ethyl]phenyl}phthalimide:
PPT-123
Mp 196.5–199.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.79 (dd, 2H, J = 5.5, 3.1 Hz), 7.35 (d, 2H,
J = 9.2 Hz), 7.32 (d, 2H, J = 8.5 Hz), 7.13 (d, 2H, J = 8.5 Hz), 6.85 (d,
2H, J = 9.2 Hz), 3.80 (s, 3H). FAB-MS m/z: 357 [M]+, 358 [M+H]+.
Anal. Calcd for C23H19NO3: C, 77.29; H, 5.36; N, 3.92. Found: C,
77.21; H, 5.51; N, 3.83.
4.1.26. N-{4-[(1E)-2-Phenylethenyl]phenyl}phthalimide: PPT-
108
Mp 297.0–299.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.97 (dd, 2H,
J = 5.5, 3.1 Hz), 7.81 (dd, 2H, J = 5.5, 3.1 Hz), 7.65 (d, 2H,
J = 8.5 Hz), 7.54 (d, 2H, J = 7.9 Hz), 7.46 (d, 2H, J = 8.5 Hz), 7.38
(t, 2H, J = 7.9 Hz), 7.28 (tt, 1H, J = 7.9, 1.2 Hz), 7.15 (s, 2H).
FAB-MS m/z: 325 [M]+, 326 [M+H]+. Anal. Calcd for
4.1.33. 4,5,6,7-Tetrachloro-N-{4-[2-(4-methoxyphenyl)ethyl]
phenyl}phthalimide: PPT-124
Mp 209.5–211.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.30 (s, 4H),
7.10 (d, 2H, J = 8.5 Hz), 6.84 (d, 2H, J = 8.5 Hz), 3.80 (s, 3H), 2.96–
2.93 (m, 2H), 2.91–2.90 (m, 2H). FAB-MS m/z: 494 [M+H]+, 495
[M+2]+, 496 [M+3]+, 497 [M+4]+, 498 [M+5]+. Anal. Calcd for
C23H15Cl4NO3: C, 55.79; H, 3.05; N, 2.83. Found: C, 55.43; H,
3.20; N, 2.76.
C
22H15NO2Á1/3H2O: C, 79.74; H, 4.77; N, 4.23. Found: C, 79.50;
H, 4.69; N, 4.19.
4.1.27. 4,5,6,7-Tetrachloro-N-{4-[(1E)-2-phenylethenyl]
phenyl}phthalimide: PPT-109
4.1.34. N-{4-[2-(3-Hydroxyphenyl)ethyl]phenyl}phthalimide:
PPT-125
Mp 296.0–296.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.65 (d, 2H,
J = 8.5 Hz), 7.54 (d, 2H, J = 7.3 Hz), 7.42 (d, 2H, J = 8.5 Hz), 7.38 (t,
2H, J = 7.3 Hz), 7.29 (tt, 1H, J = 7.3, 1.2 Hz), 7.17 (d, 1H,
J = 16.5 Hz), 7.13 (d, 1H, J = 16.5 Hz). FAB-MS m/z: 461 [M]+, 462
[M+H]+, 463 [M+2]+, 464 [M+3]+, 465 [M+4]+, 466 [M+5]+. Anal.
Calcd for C22H11Cl4NO2: C, 57.05; H, 2.39; N, 3.02. Found: C,
56.83; H, 2.69; N, 3.09.
Mp 229.0–232.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.80 (dd, 2H, J = 5.5, 3.1 Hz), 7.35–7.30 (m, 4H),
7.17 (t, 1H, J = 7.9 Hz), 6.80 (d, 1H, J = 7.3 Hz), 6.68 (dd, 1H,
J = 7.9, 2.4 Hz), 6.60 (s, 1H), 2.98–2.95 (m, 2H), 2.92–2.89 (m, 2H).
FAB-MS m/z: 343 [M]+, 344 [M+H]+. Anal. Calcd for C22H17NO3Á1/
5H2O: C, 76.15; H, 5.05; N, 4.04. Found: C, 76.38; H, 5.24; N, 4.05.
4.1.28. 4,5,6,7-Tetrachloro-N-{4-[(1E)-2-(3,4-dimethoxyphenyl)
ethenyl]phenyl}phthalimide: PPT-112
4.1.35. 4,5,6,7-Tetrachloro-N-{4-[2-(3-hydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-136
Mp 281.0–284.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.62 (d, 2H,
J = 8.5 Hz), 7.40 (d, 2H, J = 8.5 Hz), 7.11 (d, 1H, J = 16.2 Hz), 7.09–
7.05 (m, 2H), 7.00 (d, 1H, J = 16.2 Hz), 6.88 (d, 1H, J = 8.5 Hz), 3.96
(s, 3H), 3.91 (s, 3H). FAB-MS m/z: 521 [M]+, 522 [M+H]+, 523
[M+2]+, 524 [M+3]+, 525 [M+4]+, 526 [M+5]+. Anal. Calcd for
Mp 267.0–268.0 °C. 1H NMR (500 MHz, DMSO-d6) d: 9.25 (br s,
1H), 7.40 (d, 2H, J = 7.9 Hz), 7.31 (d, 2H, J = 7.9 Hz), 7.06 (t, 1H,
J = 7.3 Hz), 6.67 (d, 1H, J = 7.3 Hz), 6.66 (s, 1H), 6.58 (d, 1H,
J = 7.3 Hz), 2.92–2.90 (m, 2H), 2.84–2.81 (m, 2H). FAB-MS m/z:
479 [M]+, 480 [M+H]+, 481 [M+2]+, 482 [M+3]+, 483 [M+4]+, 484
[M+5]+. HRMS (FAB) calcd for C22H13Cl4NO3 478.9650; found:
478.9694 (M)+.
C24H15Cl4NO4Á1/2H2O: C, 54.16; H, 3.03; N, 2.63. Found: C, 54.24;
H, 2.97; N, 2.55.
4.1.29. N-{4-[(1E)-2-(3,4-Dihydroxyphenyl)ethenyl]phenyl}
phthalimide: PPT-113
4.1.36. N-{4-[2-(4-Hydroxyphenyl)ethyl]phenyl}phthalimide:
PPT-137
Mp 278.0–281.0 °C. 1H NMR (500 MHz, DMSO-d6) d: 7.97 (dd,
2H, J = 5.5, 3.1 Hz), 7.90 (dd, 2H, J = 5.5, 3.1 Hz), 7.67 (d, 2H,
J = 8.5 Hz), 7.40 (d, 2H, J = 8.5 Hz), 7.14 (d, 1H, J = 16.2 Hz), 7.01
(d, 1H, J = 2.4 Hz), 6.97 (d, 1H, J = 16.2 Hz), 6.89 (dd, 1H, J = 8.5,
1.8 Hz), 6.73 (d, 1H, J = 7.9 Hz). FAB-MS m/z: 357 [M]+, 358
[M+H]+. Anal. Calcd for C22H15NO4Á1/4H2O: C, 73.02; H, 4.32; N,
3.87. Found: C, 73.21; H, 4.36; N, 3.91.
Mp 285.0–288.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.80 (dd, 2H, J = 5.5, 3.1 Hz), 7.35–7.30 (m, 4H),
7.08 (d, 2H, J = 7.9 Hz), 6.77 (d, 2H, J = 8.5 Hz), 2.94–2.92 (m, 2H),
2.90–2.88 (m, 2H). FAB-MS m/z: 343 [M]+, 344 [M+H]+. Anal. Calcd
for C22H17NO3Á1/2H2O: C, 74.99; H, 5.15; N, 3.97. Found: C, 74.97;
H, 5.20; N, 3.89.
4.1.37. 4,5,6,7-Tetrachloro-N-{4-[2-(4-hydroxyphenyl)
ethyl]phenyl}phthalimide: PPT-138
4.1.30. N-{4-[2-(3-Methoxyphenyl)ethyl]phenyl}phthalimide:
PPT-121
Mp 266.0–268.0 °C. 1H NMR (500 MHz, DMSO-d6) d: 9.15 (br s,
1H), 7.37 (d, 2H, J = 8.5 Hz), 7.30 (d, 2H, J = 8.5 Hz), 7.03 (d, 2H,
J = 8.5 Hz), 6.66 (d, 2H, J = 8.5 Hz), 2.89–2.86 (m, 2H), 2.81–2.78
(m, 2H). FAB-MS m/z: 479 [M]+, 480 [M+H]+, 481 [M+2]+, 482
[M+3]+, 483 [M+4]+, 484 [M+5]+. Anal. Calcd for C22H13Cl4NO3: C,
54.92; H, 2.72; N, 2.91. Found: C, 54.60; H, 3.00; N, 2.82.
Mp 127.0–127.5 °C. 1H NMR (500 MHz, CDCl3) d: 7.96 (dd, 2H,
J = 5.5, 3.1 Hz), 7.79 (dd, 2H, J = 5.5, 3.1 Hz), 7.36–7.32 (m, 4H),
7.22 (t, 1H, J = 7.9 Hz), 6.82 (d, 1H, J = 7.3 Hz), 6.76 (dd, 1H,
J = 7.3, 1.8 Hz), 6.74 (d, 1H, J = 1.8 Hz), 3.79 (s, 3H), 3.00–2.96 (m,
2H), 2.95–2.91 (m, 2H). FAB-MS m/z: 357 [M]+, 358 [M+H]+. Anal.
Calcd for C23H19NO3Á1/4H2O: C, 76.33; H, 5.43; N, 3.87. Found: C,
76.34; H, 5.40; N, 3.87.
4.2. Expression and purification of the PA endonuclease and
PB2 627 domain proteins
4.1.31. 4,5,6,7-Tetrachloro-N-{4-[2-(3-methoxyphenyl)
ethyl]phenyl}phthalimide: PPT-122
The influenza (A/PR/8/34) H1N1 RNA polymerase PA and PB2
plasmids, pBMSA-PA and pBMSA-PB2, were sourced from the
DNA Bank, Riken BioResource Center (Tsukuba, Japan; originally
deposited by Dr. Susumu Nakada).35 The cDNA fragment corre-
sponding to the PA N-terminal endonuclease domain (residues
Mp 170.5–172.0 °C. 1H NMR (500 MHz, CDCl3) d: 7.34–7.29 (m,
4H), 7.22 (t, 1H, J = 7.9 Hz), 6.80 (d, 1H, J = 7.3 Hz), 6.76 (dd, 1H,
J = 7.9, 1.8 Hz), 6.73 (s, 1H), 3.79 (s, 3H), 3.00–2.97 (m, 2H), 2.95–
2.91 (m, 2H). FAB-MS m/z: 493 [M]+, 494 [M+H]+, 495 [M+2]+,

Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
5389
1–220) was amplified by PCR36 from pBMSA-PA using the primers
PA endonuclease forward NdeI, GCCGTTCATATGGAAGATTTTG
TGCGACAA and PA endonuclease reverse BamHI, GCCGTTGGATCCT
ATTGGTCGGCAAGCTTGCG. A cDNA fragment of the PB2 627
domain (residues 535–759, previously denoted as the PB2 3/3 do-
main28) was amplified by PCR from pBMSA-PB2 using the follow-
ing primers: PB2 Met 627, GCCGTTCATATGATGTGGGAGATTAA
TGGT and PB2 stop BamHI, GCCGTTGGATCCTTAATTGATGGCCA
TCCGAAT. These two amplified products were then subcloned into
the pET28a(+) plasmid (Novagen, Madison, WI) at the NdeI and
BamHI restriction sites. The two resulting constructs were then
introduced into BL21-CodonPlus (Stratagene, La Jolla, CA) Esche-
richia coli cells. The induction of 6x his-tagged recombinant protein
expression from these constructs were achieved by the addition of
blue black, 0.1% sodium acetate, and 9% acetic acid) and the
OD630 was measured.40 Cell viability was calculated based on a cal-
ibration of the OD630 values observed in mock-infected and virus-
only wells as 100% and 0%, respectively.
Acknowledgments
The influenza virus (A/PR/8/34) RNA polymerase PA and PB2
plasmids, pBMSA-PA and pBMSA-PB2, were provided by the DNA
Bank, RIKEN BioResource Center (Tsukuba, Japan; originally depos-
ited by Dr. Susumu Nakada) with the support of National Bio-Re-
sources Project of the Ministry of Education, Culture, Sports,
Science and Technology, Japan (MEXT). This work was supported
in part by
a Grant-in-Aid for Young Scientists (Start-up)
isopropyl b-D
-thiogalactopyranoside (IPTG)37 to TBG-M9 medium
20870037 (D.H.) from the Japan Society for the Promotion of Sci-
ence (JSPS), Japan.
and this was followed by purification using Ni2+-agarose.38 The re-
combinant PA endonuclease domain protein was further purified
to near homogeneity using a HiTrap™ Q FF column (GE Healthcare,
Buckinghamshire, UK) with the Akta™ prime plus system (GE
Healthcare). For further purification of the PB2 627 domain, the
his-tagged proteins were cleaved by thrombin and purified using
a HiTrap™ CM FF column (GE Healthcare) also with the Akta™
prime plus system.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.05.035.
References and notes
4.3. PA endonuclease activity assay
1. Bartlett, J. B.; Dredge, K.; Dalgleish, A. G. Nat. Rev. Cancer 2004, 4, 314.
2. Hashimoto, Y. Bioorg. Med. Chem. 2002, 10, 461.
3. Hashimoto, Y. Curr. Med. Chem. 1998, 5, 163.
Influenza A RNA polymerase PA endonuclease activity assays
were performed essentially as described by Dias et al.20–22 with
some modifications. Briefly, we modified the pH conditions (from
4. Hideshima, T.; Chauhan, D.; Shima, Y.; Raje, N.; Davies, F. E.; Tai, Y. T.; Treon, S.
P.; Lin, B.; Schlossman, R. L.; Richardson, P.; Muller, G.; Stirling, D. I.; Anderson,
K. C. Blood 2000, 96, 2943.
8.0 to 7.3) and used 1
phage DNA as the assay substrate. We added 0.35
nant N-terminal endonuclease domain of the PA subunit to
100 l of assay buffer in each reaction (the final concentration of
the protein was about 0.1 M). PPT analogs were then added to
l
g of M13mp18 single stranded circular
5. Hashimoto, Y. Arch. Pharm. Weinheim 2008, 341, 536.
6. Hashimoto, Y. Mini-Rev. Med. Chem. 2002, 2, 543.
7. Hashimoto, Y.; Tanatani, A.; Nagasawa, K.; Miyachi, H. Drugs Future 2004, 29,
383.
8. Taubenberger, J. K.; Reid, A. H.; Lourens, R. M.; Wang, R.; Jin, G.; Fanning, T. G.
Nature 2005, 437, 889.
lg of recombi-
l
l
9. Morse, S. S. Nat. Med. 2007, 13, 681.
the reaction and products were analyzed by agarose electrophore-
sis and stained with ethidium bromide.
10. Horimoto, T.; Kawaoka, Y. Nat. Rev. Microbiol. 2005, 3, 591.
11. Hatta, M.; Gao, P.; Halfmann, P.; Kawaoka, Y. Science 2001, 293, 1840.
12. Neumann, G.; Noda, T.; Kawaoka, Y. Nature 2009, 459, 931.
13. De Clercq, E. Nat. Rev. Drug Disc. 2006, 5, 1015.
14. Hayden, F. G.; Atmar, R. L.; Schilling, M.; Johnson, C.; Poretz, D.; Paar, D.; Huson,
L.; Ward, P.; Mills, R. G. N. N. Eng. J. Med. 1999, 341, 1336.
15. Reece, P. A. J. Med. Virol. 2007, 79, 1577.
16. Collins, P. J.; Haire, L. F.; Lin, Y. P.; Liu, J.; Russell, R. J.; Walker, P. A.; Skehel, J. J.;
Martin, S. R.; Hay, A. J.; Gamblin, S. J. Nature 2008, 453, 1258.
17. Izumi, Y.; Tokuda, K.; O’dell, K. A.; Zorumski, C. F.; Narahashi, T. Neurosci. Lett.
2007, 426, 54.
18. Honda, A.; Ishihama, A. Biol. Chem. 1997, 378, 483.
19. Honda, A.; Mizumoto, K.; Ishihama, A. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
13166.
20. Yuan, P.; Bartlam, M.; Lou, Z.; Chen, S.; Zhou, J.; He, X.; Lv, Z.; Ge, R.; Li, X.; Deng,
T.; Fodor, E.; Rao, Z.; Liu, Y. Nature 2009, 458, 909.
21. Dias, A.; Bouvier, D.; Crépin, T.; McCarthy, A. A.; Hart, D. J.; Baudin, F.; Cusack,
S.; Ruigrok, R. W. Nature 2009, 458, 914.
22. Kuzuhara, T.; Iwai, Y.; Takahashi, H.; Hatakeyama, D.; Echigo, N. PLoS Curr.
Influenza 2009. RRN1052.
23. Subbarao, E. K.; London, W.; Murphy, B. R. J. Virol. 1993, 67, 1761.
24. Shinya, K.; Hamm, S.; Hatta, M.; Ito, H.; Ito, T.; Kawaoka, Y. Virology 2004, 320,
258.
4.4. Electrospray ionization (ESI) mass spectrometry (MS)
Chemicals at a dose of 50
protein at 1–3 M were mixed in 200
tate/methanol (1:1), to which 1% acetic acid was added. This mix-
ture was then injected at 100
l/h into an ESI-MS.30,31 For binding
l
M and recombinant PB2 627 domain
l
ll of 10 mM ammonium ace-
l
analysis, recombinant PB2 627 domain protein was used.28 Mea-
surements were performed with a Fourier transform ion cyclotron
resonance mass spectrometer (FT-ICRMS, Bruker Daltonics),
equipped with an ESI ion source, using the following parameters:
capillary À4.0 kV, spray shield À3.5 kV, dry temp 40 °C, resolution
at m/z 2792 50 k, flow rate 100 ll/h, solvent 10 mM ammonium
acetate buffer 49.5%, methanol 49.5%, and acetic acid 1% (v/v/v).
4.5. Inhibition of viral growth
25. Labadie, K.; Dos Santos Afonso, E.; Rameix-Welti, M. A.; van der Werf, S.;
Naffakh, N. Virology 2007, 362, 271.
26. Gabriel, G.; Dauber, B.; Wolff, T.; Planz, O.; Klenk, H. D.; Stech, J. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 18590.
27. Tarendeau, F.; Crepin, T.; Guilligay, D.; Ruigrok, R. W.; Cusack, S.; Hart, D. J. PLoS
Pathog. 2008, 4, e1000136.
28. Kuzuhara, T.; Kise, D.; Yoshida, H.; Horita, T.; Murazaki, Y.; Nishimura, A.;
Echigo, N.; Utsunomiya, H.; Tsuge, H. J. Biol. Chem. 2009, 284, 6855.
29. Motoshima, K.; Ishikawa, M.; Sugita, K.; Hashimoto, Y. Biol. Pharm. Bull. 2009,
32, 1618.
Madin–Darby canine kidney (MDCK) cells39 were cultured in
MEM (Minimum Essential Medium; Gibco/Invitrogen, Carlsbad,
CA) supplemented with 10% fetal bovine serum in 5% CO2 incuba-
tor at 37 °C. A confluent monolayer of MDCK cells was prepared in
each well of a 96-well plate. Various concentrations (0.63–80 lM)
of PPT analogs were mixed with or without 100 TCID50 (50% of the
infectious dose) of H1N1 influenza A virus (A/Puerto Rico/8/34
(PR8)) in the presence of trypsin and incubated at 37 °C for
30 min.40 MDCK cells were washed with PBS(À) and the viral mix-
ture was added to the cells. Treated cells were then incubated for
four days at 34 °C under 5% CO2. After incubation, the medium
was removed and cells were fixed with a 10% formaldehyde solu-
tion. Viable cells were stained with NB solution (0.1% naphthol
30. Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989,
246, 64.
31. Baca, M.; Kent, S. B. H. J. Am. Chem. Soc. 1992, 114, 3992.
32. Yang, C. S.; Wang, X.; Lu, G.; Picinich, S. C. Nat. Rev. Cancer 2009, 9, 429.
33. Song, J. M.; Lee, K. H.; Seong, B. L. Antiviral Res. 2005, 68, 66.
34. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.;
Olson, A. J. J. Comput. Chem. 1998, 19, 1639.
35. Nakamura, Y.; Oda, K.; Nakada, S. J. Biochem. 1991, 110, 395.

5390
Y. Iwai et al. / Bioorg. Med. Chem. 18 (2010) 5379–5390
36. Mullis, K. B.; Faloona, F. A. Methods Enzymol. 1987, 155, 335.
37. Studier, F. W.; Moffatt, B. A. J. Mol. Biol. 1986, 189, 113.
38. Janknecht, R.; de Martynoff, G.; Lou, J.; Hipskind, R. A.; Nordheim, A.;
Stunnenberg, H. G. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 8972.
39. Taub, M.; Saier, M. H., Jr. Methods Enzymol. 1979, 58, 552.
40. Hierholzer, J. C.; Killington, R. A. Virology Methods Manual; Mahy, B. W. J.,
Kangro, H. O., Eds.; Academic Press Limited: London, 1996; Chaper 2, pp 35–37.