https://www.abbs.info     e-mail: [email protected]
ISSN 0582-9879                                    Acta Biochim Biophys Sin 2004, 36(1): 1–10                                      CN 31-1300/Q

Understanding
SARS with the New Kind of Science

Da-Wei LI1,2, Yu-Xi PAN1,2, Yun DUAN2,1, Zhen-De
HUNG1, Ming-Qing XU2,1, and Lin HE2,1*

(1Bio-X
Life Science Research Center, Shanghai Jiao Tong University, Shanghai 200030,
China;2Institute for Nutritional science, the
Chinese Academy of Sciences, Shanghai 200031, China )

Abstract
Stepping acquired immunodeficiency syndrome (AIDS),
severe acute respiratory syndrome (SARS) as another type of disease has been
threatening mankind since late last year. Many scientists worldwide are making
great efforts to study the etiology of this disease with different approaches.
13 species of SARS virus have been sequenced. However, most people still
largely rely on the traditional methods with some disadvantages. In this work,
we used Wolfram approach to study the relationship among SARS viruses and
between SARS viruses and other types of viruses, the effect of variations on
the whole genome and the advantages in the analysis of SARS based on this novel
approach. As a result, the similarities between SARS viruses and other coronaviruses
are not really higher than those between SARS viruses and non-coronaviruses.

Key
words genome sequence; SARS; visualization; Wolfram approach

In this work, we tried to understand the
pathogenesis of SARS, the world’s threat [1–14] using a complete novel approach
[15–20], or Wolfram approach which was systematically described in the book
entitled “A New Kind of Science” in 2001 and has drawn extensive
attention in the world [15]. In contrast with the traditional methods of DNA
sequence comparison, Wolfram approach was based on the concept that simple
rules are able to produce highly complicated behaviour such as dynamically  viewing alterations on visualization,
including transposition, insertion, deletion and duplication, etc., in a whole
genome scale, and even in a single base scale when the base precisely located.
Furthermore, it has become possible to make progress on a remarkable range of
fundamental issues of lives that have never been successfully studied by any of
the existing sciences based on traditional mathematical rules, which are
limited in exploring the complex behavior in a typical biological system. For
example, the evolutionary theory cannot really or completely explain the origin
of complexity of biological system [15]. Research and speculation in living
organisms at a molecular level that was normally neglected by Wolfram, have
little success for the explanation of complexity in lives.

With
Wolfram approach-based method, we explored both the simple rules and a special
rule from the 256 rules suggested by Wolfram [15] to avoid traditional
intuition that the behaviour must be simple if the rule for a program is
simple. This is not true from the data demonstrated by both Wolfram’s work and
our own work. The remarkably simple rule can actually capture the essential
mechanisms responsible for complex phenomena in living organisms.

In
order to gain an insight into SARS, we analyzed DNA sequences of different
viruses in detail by the simple rules, initial conditions and highly complex
behaviour of the final images were studied visually.

Materials
and Methods

Sequences
of SARS viruses and other related viruses

All studied sequences including 13 SARS
viruses were downloaded from free database of National Center for Biotechnology
Information (NCBI):

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=nucleotide&cmd=search&term=SARS.

The genomes of SARS viruses in Fig. 1 are as
follows:

SARS
BJ01, partial genome;

SARS
BJ02, partial genome;

SARS
BJ03, partial genome;

SARS
BJ04, partial genome;

SARS
CUHK-W1, complete genome;

SARS
GZ01, partial genome;

SARS
HKU-39849, complete genome;

SARS
TOR2, complete genome;

SARS
Urbani, complete genome;

SARS
coronavirus CUHK-Su10, complete genome;

SARS
coronavirus isolate SIN2774 complete genome;

SARS
coronavirus TW1, complete genome;

SARS
coronavirus, complete genome.

 

Fig.
1       
Viewing SARS viruses on the evolutionary trees

(A) HIV
represented by the Arabic numerals. (B) All 13 SARS viruses. (C) A new tree
formed after adding 10 coronaviruses represented by Arabic numerals
corresponding to the number in front of species appeared in d and 6 murine
hepatitis viruses represented by M in the black frame on the basis of the tree
of 13 SARS viruses. The contents in blue frame correspond to those in blue
frame in (D), which including SARS-CUHKW1 and other 7 SARS viruses are very
stable on the position of the trees after the samples have been added over and
over but another five SARS species including SARS-GZ framed in red are greatly
variable. (D) The complicated relationship shown

on the
final tree composed of viruses including 13 SARS viruses, 16 HIV and 43 others.

The HIV strains in Fig. 1(A) are as follows:

1,
HIV-1 strain CNGL179 from China;

2,
Human immunodeficiency virus 1;

3,
HIV-1 isolate BK132 from Thailand;

4,
HIV-1 isolate CNHN24 subtype B-Thai from China;

5,
HIV-1 isolate 01IN565.14 from India;

6,
HIV-1 isolate 96GH2911;

7,
HIV-1 isolate 95SN7808 from Senegal;

8,
HIV-1 isolate BZ163 from Brazil;

9,
HIV-1 strain 98IS002 from Israel;

10,
HIV-1 strain TV012 clone 2-1_4 subtype C from South Africa;

11,
HIV-1 patient WCIPR sample 1990 clone 32;

12,
HIV-1;

13,
HIV-1 isolate X397 from Spain;

14,
HIV-1 strain 98ZA528 clone 6 from South Africa;

15,
HIV-1 isolate US4 from USA;

16,
HIV-1 isolate 00BW3891.6 from Botswana.

Basic
principle of Wolfram approach and the rule selected in the study

DNA sequence is currently thought to be one
dimensional linear structure coded in two states: 0 and 1 (black and white seen
in images), each cell in the coding sequence has immediate interaction with its
neighbouring cells at left and right, respectively, in terms of “positive”
power and “negative” power. The powers from either side, which are not always
equal, can be scored between 0 and 1 representing for or against each other.
This is the mode of power relation of one dimension sequence [Fig. 2(A)]. In
this mode, only eight arrangements can be found [15] [Fig. 2(B)]. The cells in
the first row and in the second row are paternal and generated, respectively
[Fig. 2(C)].

 

Fig.
2       
Eight arrangements of 0 and 1 in the mode of power

(A) The
mode of power relation of one dimension sequence. (B) In the mode of power
interaction with eight arrangements. (C) The cells in the first row and in the
second row are paternal and generated, respectively, suggesting the simple
rules we selected in our study.

 

If the cell and its two neighbours are all in
state 0, which can be seen on the right of Fig. 2, the cell interacted with the
same kind of powers from both sides keeps the original state 0 in the next
step. In 001, the middle cell accepting unequal powers, which are opposite each
other from both sides, still maintains the final state 0 in the next step with
the two powers overlapped.

However,
in 010, the powers which belong to the same kind of powers from both sides are
greater than the power of the cell itself, resulting in the state 1 shifted to
state 0 in the next step. By analogy, 011 will be changed to state 1, 101 to
state 1, 111 remains the same state 1.

However,
in 100 and 110, of which both have neighbours 1 on the left and neighbour 0 on
the right, the final state in the next step is the opposite of the initial
state. All the eight arrangements result in very complicated behaviours of DNA
sequence with two kinds of lines and the nested structure. In theory, still 255
more rules exist but the present rule provides the best solution.

Computer
and image

The computer Origin 3000 (Silicon Graphics,
Inc. 64 500 MHZ IP35 processors) was used throughout this study. Each sequence
was run on the same simple programs to see how they behave. Eventually, one of
the rules was selected for comparative study based on more than 3,000 images
(200,000 Mega) showing behaviours of complexity.

Results
and Discussion

Characteristics
of SARS viruses

In the study of sequence homology [21–24], we
compared 62 viruses including 13 SARS viruses, 16 HIV and some related viruses
on the evolutionary tree, which was constructed by comparing the sequence
homology between any two viruses [24–28] using Clustal (Fig. 1). Bovine
coronavirus and avian infectious bronchitis virus, as SARSrelated viruses,
showed higher score. Previous studies [21–28] reported that SARS virus was a
new kind of coronavirus similar to bovine coronavirus and avian infectious
bronchitis virus. However, our work indicated that they had remarkable
intrinsic differences.

All the 13 SARS viruses studied behaved quite
differently (see Fig. 3). There was a very large nested structure across the
beginning 10 kb regions that was rarely seen in all other studied viruses
except those with higher homology. Furthermore, four smaller nested structures
(approx. 2 kb long on average) were found within approx. 12–25 kb in SARS
viruses but not in most other studied viruses (Fig. 4). Therefore, it is
rational to think that the nested structure is the typical feature of SARS
viruses containing some special bio-information absent in other viruses, which
may be involved in the development of SARS virus by producing replicase or
vital protein in this region as suggested previously [21] (Fig. 5).

 

Fig.
3       
Images of all 13 SARS virus

The images are
arranged according to the colour order in chromatogram. The more similar the
colours are, the closer their consanguinity should be, suggesting the degree of
similarity in behaviours of SARS viruses.

Fig.
4       
Comparison of images among five different viruses

(A) and
(B), SARS virus and equine rhinovirus, respectively, showing clear the nested
structure with probably closer relationship. (C) Another virus in which the
nested structures were also found but the relationship with SARS virus is not
as close as that between SARS virus and equine rhinovirus. (D) A typical
behaviour of a common virus. (E) A behaviour of HIV. Note: Images in (A), (C),
(D), and in (B), (E) are reduced with 15,000 and 1600 folds, respectively.

Fig. 5        The nested structure and
both Replicases 1A and 1B of SARS viruses found in the same region.

By
comparison, we found the nested structure are mainly located in the beginning
21.5 region of SARS virus sequence, which is just in the position for both Replicases
Fig. 1(A) and (B), an important “weapon” for SARS viruses to maintain their
amounts necessarily for causing the disease.

Pair wise comparison in the traditional way
also exhibited similarity of sequence among porcine transmissible gastroen, human
coronavirus 229E, and SARS viruses but no fundamental similarity was found in
equine rhinovirus (Fig. 1). However, the whole genome of equine rhinovirus (ER)
had similar regulating feature of SARS virus (Fig. 4), although ER genome
contained only 7734 bp as compared to more than 29 kb of SARS genome [21–24].
Using traditional method, SARS virus was found to have high homology with
porcine transmissible gastroen and human coronavirus 229E, but we found that
SARS virus and equine rhinovirus had similar intrinsic image and specific
nested structure (Fig. 4). ER genome is much smaller than SARS virus genome, so
it is unlikely that ER was mutated directly to SARS virus but we could not
exclude the possibility that ER is the ancestor of SARS virus. At the same time,
SARS virus and human coronavirus 229E are very different in behaviour. Another
interesting finding is that the beginning genome sequences after the first
“atgccc” of other coronaviruses have very similar behaviour. Furthermore, the
behaviours of the last 2 kb region of each species are also similar.

Possible
origin of SARS virus

SARS-GZ was likely to be formed on the basis
of SARSCUHK because its 33 bp in the beginning region, about 300 bp in the
ending region [including 24 poly (A)] and 520 bp between 21,300 and 21,820 bp
do not exist in SARS-GZ. However, seven pieces of 7–13 bp showing 100%
homology, and dozen pieces of 10-30 bp showing 70%–99% homology, certain
sequences in the 520 bp in SARS-CUHK virus DNA were found to be inserted into
the regions within 3.7–4.2 kb or 26.5–26.7 kb, and at least in 6 other
positions in SARS-GZ genome (Fig. 6).

 

Fig.
6 Comparisons between SARS-CUHK and SARS-GZ

In the upper three images, the green
colour stands for the entire genome of SARS-CUHK and the red colour for the
genome of SARS-GZ. (A) In SARS-GZ (red) there is a 520 bp of deletion indicated
by the gap but other parts matched well. (B) and (C), In the second and third
images with partial magnification show small pieces of DNA inserted into at
least five positions of SARS-GZ (red). (D) The tiny difference at base level
can be detected from the magnified original images, the one in green stands for
SARS-CUHK.

Thus, the first case of SARS
reported in Guang Dong province was probably caused by SARS virus originated in
Hong Kong. The evolutionary position of SARS-CUHK was found to be more stable
than that of SARS-GZ, supporting the view that SARS-CUHK appeared earlier than
SARSGZ. Furthermore, the eight SARS viruses other than SARSGZ, -BJ03, -BJ04,
-SIN2774, and -Coronavirus are more stable. The three strains from Hong Kong
belong to three different subgroups, this fact also supports the above view
(Fig. 1).

Advantages
in the study of SARS with Wolfram approach-based method

The array in Fig. 7 is only consisted of two
elements, white and black cells (the smallest unit in coding sequence),
representing power and counter-power. The first (original) line composed of the
cells in the array refers to the coding sequence of DNA. Under the same rule,
every cell changes its state in each following step. If one cell is replaced
with another cell, for example, after a white cell turns to a black cell, an
“earthquake” in the following steps can take place. In the example of p53 (Fig.
8), only one base change (A → T at 186 bp position) results in the continuing abnormal
expression of genetic information (white lines) which should be terminated. On
the other hand, other genetic information, which should not be expressed in the
late stage of life development, is continually expressed. Hence, despite that
high similarity of sequences in a certain region between two species can
sometimes be found, it does not necessarily mean that their final behaviours
should be the same. Instead, they can be greatly different. This may be
explained by the fact that the similar sequence is also regulated by the omplex
network of power interactions among the whole genome.

 

Fig.
7 The power and counter-power shown with white cell or white line and black
cell or black line, respectively

(A)
Interaction between “positive” and “negative” powers. It needs a rule installed
for gaining a final balanced power. (B) The power in black beats the power in
white, which is partially driven away. (C) and (D), The biological system is in
a balanced state under the circumstance the power in white does not exist any
more. (E) While interaction between the power in while and power in black is
equivalent in local area, a “completely balanced state” has been reached but
can only be formed in partial area in genetic material. (F) The crossover of
“positive” and “negative” powers at genome level no longer exists after the
state balanced to finally direct the trend of development of organism.

On the contrary, sequences obtained in the
traditional way show less similarity but the distribution of interaction powers
of bases in the network of power interactions among the whole genome can be
very similar in behaviours, such as the process of pathogenesis between the two
species. From this work, we think that SARS viruses are likely to catch a part
of genetic information from other non-coronaviruses to increase their special
power of causing the disease as a result of the partial similarities of
behaviours between SARS and other non-coronaviruses like HIV.

In the images of power (Fig. 7),
the thick and thin lines have been divided into two types, the white one
inclining towards left and the black one towards right with 45 degree angle. If
the equal numbers of white and black lines meet each other, it will result in
an interim balanced state between the two powers (Fig. 7). However, in most
circumstances the two lines eventually disappear because of the continence of
the powers. While the number of black lines is much more than the number of
white lines, the power represented by black lines will remain because of its
dominant role. The remaining power, which is approximately equal to the power
after the white line has been deducted from the black line, gets weaker, and
vice versa. Nevertheless, the white line or the black line will not always go on
if the counter-powers meet again in the following step. Both lines will
disappear to reach a balanced state if they are equal. Otherwise, the line with
more power will dominate the situation with the deducted power. Through their
“encounter” again and again, at last only one type of line will exist with a
final stably balanced state. Based on the interaction of power and
counter-power, the “encounter” between the powers in white towards the left and
the power in black towards the right in several locations may finally result in
the formation of the nested structure, which is presumed to be the important
state of life filled with complex regulation network.

In analysis of the whole genome behaviour
with 10,000 fold reduction of the original size, we found that behaviours of
all studied organisms can be classified into three categories: the first one
mainly composed of white lines inclining towards left in 45 degree as most
common viruses show; the second one mainly composed of black lines towards
right in 45 degree as HIV exhibits; and the last one composed of the typical
nested structure as SARS viruses possess. Interestingly, in this study the
three kinds of viruses happen to have these three types of behaviours (Fig. 4),
suggesting that the behaviours of HIV and SARS are specific and rare. The
typical feature of Wolfram approach is the conversion of “A” “G” “C” “T” into
two states, white and black, in accordance with the characteristics that
opposite sides coexist in all substances and also with the theory of “Yin-Yang”
balance (Fig. 7).

All genetic materials are made up of
thousands and hundreds of bases of AGCT or AGCU, but there must be an inner
uncertain mode of base organization that each species follows in the
utilization of these four bases, so different viruses show different behaviours
even under the same rule. Though sometimes the result obtained by both the
traditional and the new methods are similar, one of the important differences
between them is due to the fact that the judgment with the traditional way
always relies on homology or the variation of a single base, etc., which we
think is not complete or objective. Instead, Wolfram approach pays attention to
the interaction power of regulation between any adjacent bases, reflecting a
major and basic power determining the development and evolution of organism
along its own specific track. In fact, the interaction power between bases can
be divided into two categories: promotion and continence with each other (Fig.
7), which can be much more effective than single base alterations in the
development of organisms. In addition, each base receives powers from
neighbouring bases and exerts power to neighbouring bases at the same time,
indicating that the role of each base in a sequence is variable. When all
powers from different bases are gathered together, a network of huge power
interactions with overlap of powers will be formed, but the balance will be
easily broken by an alteration of sequence occurs because of its precision and
fragility. As we know, sequence alteration sometimes can be ignored, but
sometimes can not because the original base may play a key role in the
behaviour of the whole genome. This also explains why some mutations do not
affect life too much but sometimes even a single base mutation can cause death,
which we call “key (vital) base unbalance” (Fig. 8). The Wolfram approach is
just a good way to answer this sort of questions.

Ignoring the power interactions between bases
in the traditional analysis usually causes unsatisfactory explanation or
misleads the direction of research, especially in the study of some complex
phenomenon of lives. At present, researchers are trying hard to produce drugs
against SARS.For example, human rhinovirus 3 Cpro inhibitors have been
considered for SARS therapy [29]. However, equine rhinovirus (ER) has more
similar regulating characteristics of SARS as shown in the analysis by Wolfram
approachbased method (Fig. 4). Therefore, ER-based inhibitor could be better
than human rhinovirus-based inhibitor for fighting SARS.

 

Fig. 8 Eight arrangements of 0 and 1 in the
mode of power

(A) The mode
of power relation of one dimension sequence. (B) In the mode of power
interaction with eight arrangements. (C) The cells in the first row and in the
second row are paternal and generated, respectively, suggesting the simple
rules we selected in our study.

In analysis with Wolfram approach, SARS virus
shows an image of much more outstanding and complicated behaviours under one of
256 rules (Fig. 8). It suggests that SARS should have a specific
bio-information with its own mode of base organization, which is not necessary
to be the same as HIV that can follow another type of mode of base organization
(Fig. 4). According to the current thinking, however, HIV, which is still a
kind of virus to cause huge troubles in man, should have a large number of similar
complex information in the image but, instead, it looks more special than what
we have expected. This may be also a reason why people are facing big
difficulties to answer many scientific questions using the traditional way,
which people have noted is actually limited in use.

Our result is different from those reported
recently in other papers [30–31] which believe SARS is one species of
coronavirus genus, but our image results shows that SARS could be a new kind of
virus, which may be far away form other coronaviruses in family-relation
distance. This can be due to the fact that traditional sequence analysis
ignores the difference between any two bases. In contrary, Wolfram approach is
sensitive even to the spot mutation of a single base, and our result shows the
inner characteristics and future trend of the whole genome. As a result, the
similarities between SARS viruses and other coronaviruses are not really higher
than those between SARS viruses and other non-coronaviruses. Even in the result
obtained by Guan et al.[32], the difference between the sequences of
SARS viruses and coronaviruses from Masked pa1m civet can be found. Therefore,
more work remains to be done to clarify the cause in the future.

Briefly,
we are the first one to apply Wolfram approach in the analysis of SARS virus at
the molecular and sequence level with many advantages, particular in the
magnification of tiny changes in DNA sequence for both detailed and overall
analysis in the whole genome scale. If the analysis is only based on the
comparison of sequences in the traditional way, the scope of research will be
limited. In this work, we studied and discussed the origin of SARS with Wolfram
approach but the etiology of SARS investigated in this study may need more work
to further confirm.

References

1
Stephenson J. Global Impact of SARS. JAMA, 2003, 289: 2349

2 Mackay B.
SARS: “A domino effect through entire system”. CMAJ, 2003, 168: 1308

3 Chandler
C. SARS attacks, China shudders. Fortune, 2003, 147: 32

4 Lemonick MD,
Park A. The truth about SARS. Time, 2003, 161: 48–53

5 Shute N.
SARS hit home. US News World Rep, 2003, 134: 38–42

6 Kondro W.
Canadians still stung by WHO’s SARS travel advisory. Lancet, 2003, 361: 1624

7
Shortridge KF. SARS exposed, pandemic influenza lurks. Lancet, 2003, 361: 1649

8 Wenzel
RP, Edmond MB. Managing SARS amidst uncertainty. N Engl J Med, 2003, 348:
1947–1948

9 Parry J.
Containment of SARS depends on how it is handled in China. BMJ, 2003, 326: 1004

10 Parry J.
WHO warns that death rate from SARS could reach 10%. BMJ, 2003, 326: 999

11 Cameron
PA, Rainer TH, de Villiers Smit P. The SARS epidemic: Lessons for Australia.
Med J Aust, 2003, 178: 478–479

12 Clark J.
Fear of SARS thwarts medical education in Toronto. BMJ, 2003, 326: 784

13 Gerberding
JL. Faster… but fast enough? Responding to the epidemic of severe acute
respiratory syndrome. N Engl J Med, 2003, 348: 2030–2031

14
Cyranoski D. Taiwan left isolated in fight against SARS. Nature, 2003, 422: 652

15 Wolfram
S. A new kind of science (Wolfram Media, Inc., Champaign, 2001)

16 Giles J.
What kind of science is this? Nature, 2002, 417: 216–218

17 Casti
JL. Science is a computer program. Nature, 2002, 417: 381–382

18 Hayes B.
The world according to Wolfram. American Scientist, 2002, 90: 308–312

19 Mitchell
M. Is the universe a universal Computer? Science, 2002, published online,

http://www.sciencemag.org/cgi/content/full/298/5591/65

20 Lemonick
ND. How everything works. Time, 2002, 159(20):67, published online, http://www.ph.sophia.ac.jp/~boes-ken/E-Wolfram-Automata.pdf

21 Marra
MA, Jones SJ, Astell CR, Holt RA, Brooks-Wilson A, Butterfield YS, Khattra J et
al. The genome sequence of the SARS-associated coronavirus. Science, 2003,
published online, http://www.sciencemag.org/cgi/rapidpdf/1085953v1.pdf

22 London
OD. Two strains of the SARS virus sequenced. BMJ, 2003, 326: 999

23 Enserink
M, Vogel G. Infectious diseases. Hungry for details, scientists zoom in on SARS
genomes. Science, 2003, 300: 715–717

24 2003,
Published online,
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=nucleotide&c
md=search& term=SARS

25 Wang Y,
Ma WL, Song YB, Xiao WW, Zhang B, Huang H, Wang HM et al. Gene sequence
analysis of SARS-associated coronavirus by nested RT-PCR. Di Yi Jun Yi Da Xue
Xue Bao, 2003, 23: 421–423

26 Rota PA,
Oberste MS, Monroe SS, Nix WA, Campagnoli R, Icenogle JP, Penaranda S et al.
Characterization of a novel coronavirus associated with severe acute
respiratory syndrome. Science, 2003. published online,
http://www.sciencemag.org/cgi/rapidpdf/1085952v1.pdf

27 Yang J,
Wang ZH, Chen JJ, Hou JL. Clinical detection of polymerase gene of
SARS-associated coronavirus. Di Yi Jun Yi Da Xue Xue Bao, 2003, 23: 424–427

28 Holmes
KV. SARS-associated coronavirus. N Engl J Med, 2003, 348: 1948– 1951

29 Anand K,
Ziebuhr J, Wadhwani P, Mesters JR, Hilgenfeld R. Coronavirus main proteinase (3CLpro)
structure: Basis for design of anti-SARS drugs. Science, 2003, published
online, http://www.sciencemag.org/cgi/rapidpdf/1085658v1.pdf

30 Qi Z, Hu
Y, Li W, Chen YJ, Zhang ZH, Sun SW, Lu HC et al. Phylogeny of SARS-CoV
as inferred from complete genome comparison. Chinese Science Bulletin, 2003,
12: 1175– 1179

31 Gao L,
Qi J, Wei HB, Sun YG, Hao BL. Molecular phylogeny of coronaviruses including
human SARS-CoV. Chinese Science Bulletin, 2003, 12: 1170–1175

32 Guan Y,
Zhang BJ, He Y, Lin XL, Zhang ZX, Cheung CL, Luo SW et al. Isolation and
characterization of viruses related to the SARS coronavirus from animals in
Southern China. Science, 2003, 302: 276–278

 

Received:
September 19, 2003 Accepted: November 4, 2003

This work
was supported by grants from the Major State Basic Research Development Program
of China(973 Program)(No. 2001CB510304), the National Technology Research and
Development Program of China Projects (863 program) (No. 2001AA227021),
Shanghai Municipal commission for Science and Technology, the National Natural
Science Foundation of China (30130250), and Qiu Shi Science and Technologies
Foundation

*Corresponding
author: Tel /Fax, 86-21-62822491; E-mail, [email protected]
or [email protected]