January 8, 2008 17:13 00039
Biomedical Engineering: Applications, Basis and Communications, Vol. 19, No. 5 (2007) 295–301
THE RECOMBINATION OF HUMAN ENTEROVIRUS 71 Tzu-Ching Shih∗ and Po-Yuan Chen†
∗Department
of Medical Radiology Technology China Medical University and Department of Radiology China Medical University Hospital †Department of Biological Science and Technology China Medical University Taichung 404, Taiwan †
[email protected] Accepted 20 September 2007
ABSTRACT In 1998, the enterovirus (EV) infections outbreak in Taiwan caused 78 fatalities. Since then, EV infections have continuously posed a threat to the public. Among the 64 serotypes of enteroviruses known to infect human, the enterovirus 71(EV71) is suspected to be the major cause for severe cases. In this study, we estimate the recombination point of enterovirus 71 vp1 by using the method of Likelihood Analysis of Recombination in DNA. The datasets of enterovirus 71 DNA sequences are available in GenBank. After careful cross validation, eight candidate sequences are chosen to advance analysis, including 2734TAI98, TW227298, 1423SIN98 and other five DNA sequences as well. Then, the construction of the phylogeny trees (neighbor-joining trees will be used in this paper) would support for recombination in EV71 virus. In these two methods, the breakpoint was found to be in similar position, demonstrating that a single recombination event occurred prior to the divergence of these two strains. Keywords: Enterovirus 71; Likelihood analysis; Phylogeny tree; Recombination.
INTRODUCTION
were significant overlaps in the biological properties of viruses in the different groups. Since 1970, newly identified serotypes have not been assigned to the above groups but, rather, have been numerically classified as enterovirus serotypes (ENV) 68 to 71.16 Enterovirus infection in humans may result in a wide range of acute symptoms involving the cardiac and skeletal muscle, central nervous system (CNS) pancreas, skin, and mucous membranes, as listed in Table 1. Poliomyelitis caused by PV has been a successful World Health Organization (WHO)-sponsored Poliomyelitis Eradication Initiative (PEI), but other enterovirus infections remain frequent and sometimes
The enteroviruses belong to the genus Picornaviridae. Serologic studies have distinguished 66 human enterovirus serotypes on the basis of an antibody neutralization test. Most infections are mild, and are considered by many to be unimportant as human pathogens and unworthy of sustained investigation.17 However, enteroviruses may also result in serious or even fatal disease. On the basis of their pathogenesis in humans and experimental animals, the enteroviruses were originally classified into four groups, polioviruses, coxsackie A viruses (CA), coxsackie B viruses (CB), and echoviruses, but it was quickly realized that there †
Corresponding author: Po-Yuan Chen, Department of Biological Science and Technology, China Medical University, Taichung 404, Taiwan, ROC. 295
January 8, 2008 17:13 00039
296
T.-C. Shih & P.-Y. Chen
Table 1. Clinical Manifestations of Enterovirus Infection (Peter et al., 1998).
Syndrome
Commonly Implicated (Serotype(s))
Asymptomatic infection
All serotypes
Paralytic poliomyelitis
PV1 to PV3 ENV70 ENV71 CAV7 PV, CBV, CAV, ECV
Aseptic meningitis/ meningoencephalitis Acute myocarditis
PV, CBV, CAV, ECV ENV71 CBV
Bornholm disease (pleurodynia)
CBV
Hand, foot, and mouth disease
CAV16 ENV71
Herpangina Exanthem
CAV, CBV, ECV CAV, CBV
Acute hemorrhagic conjunctivitis
ENV70 CAV 24(variant)
Neonatal multisystem disease
CBV, ECV
Nonspecific febrile respiratory illness
CAV, CBV, ECV
causes serious morbidity. Enterovirus infection has also been associated with such chronic disease as dilated cardiomyopathy and chronic myocarditis,3,6,7 chronic fatigue syndrome,2,8 insulin-dependent diabetes mellitus,4,5 motor neuron disease,23 and postpoliomyelitis syndrome.20 Evidence from studies in murine models indicates that chronic enterovirus infection is characterized by restricted genome replication and gene expression, although some controversy about the role of enterovirus infection in chronic disease in humans remains. Chronic infections also occur in immunodeficient patients. Enterovirus 71 (EV71) is a positive-stranded RNA virus belonging to the enterovirus genus of the Picornaviridae family. It is transmitted from person to person mainly by the fecal-oral route. After replication in the mucosal system, the virus may enter the circulation (viremia) and finally find its way to the central nervous system.14 The clinical manifestations caused by EV71 infection vary from mild hand, foot and mouth disease, even fatal damage to life. In 1998, an epidemic of EV71 infection affected more than 90,000 children in Taiwan and caused 78 deaths.11,15 There is still no vaccine or antiviral drug available against this infection. In dealing with the phylogeny study, an understanding of the structure and function of the enterovirus
genome is required. The enterovirus genome is approximately 7,500 nucleotides long, of positive polarity. An approximately 750-nucleotide 5 untranslated region is followed by a long open reading frame coding for an approximately 2,100-amino-acid polyprotein. This is followed by a short 3 untranslated region and a poly(A) tail.1 The open reading following the 5 UTR is translated into a polyprotein which is co- and post-translationally cleaved to give four structural proteins (VP4, VP2, VP3, and VP1), which form the viral capsid, and seven nonstructural proteins (P2A, P2B, P2C, P3A, P3B, P3C, and P3D).12 VP1 and VP3 are partially exposed on the virion surface, while VP4 is completely internalized in infection virons. Protein 2A is one of the viral proteinase that cleaves the polyprotein in trans between proteins VP1 and 2A and frees the capsid protein precursor from the rest of the polyprotein.22 The specific functions of 2B and 2C are not known, although protein 2C and its precursor form 2BC have been found in the replication complex of PV, and protein 2C has a helicase activity. Protein 3AB is a precursor of 3B, the small polypeptide covalently linked to the 5 UTR of picornavirus RNA molecules. Protein 3C is the second viral protease, which does most of the RNA-dependent RNA polymerase. Among the 11 regions of the EV whole genome, VP1 is the most external and immunodominant of the picornavirus capsid proteins. A number of major neutralization sites reside in the VP1 proteins of many piconaviruses.18 To identify the serotypes of enteroviruses efficiently and correctly is a first important work. The classification of enterovirus 71 from other serotypes such as coxackie virus A16 by RT-PCR, and a more comprehensive pathological, virological, and molecular study were studied by Jing-Jou Yan et al.24,25 In this study, full VP1 sequences are used to examine the recombination of enterovirus 71. In the first part of this analysis, Likelihood Analysis of Recombination in DNA26 combined with HKY model10 are used to find out the most possible points of recombination in full VP1 sequences. Then, the neighbor-joining trees (with associate bootstrap values) will be supporting the phenomena of recombinations.
MATERIALS AND METHODS Enterovirus Sequences from GenBank It is started by deriving a selected set of non-redundant sequences from March 2003 release of the NCBI GenBank database of known DNA sequences. 50 sequences from NCBI GenBank (http://www.ncbi.nlm.nih.gov/
January 8, 2008 17:13 00039
The Recombination of Human Enterovirus 71 Table 2. No. 1 2 3 4 5 6 7 8 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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
297
Enterovirus71 VP1 sequences used in recombination analysis.
Strain TW227298 TW208698 NCKU98 1245aTW98 1423SIN98 1424SIN97 2731TAI98 2732TAI98 2733TAI98 2734TAI98 2735ATAI98 2785TAI98 2848TAI98 2849TAI98 2850TAI98 2851TAI98 2852TAI98 2861TAI98 2862TAI98 2864TAI98 2735BTAI98 2867TAI98 2868TAI98 2871TAI98 2782TAI98 2873TAI98 2878TAI98 2885TAI98 2892TAI98 2894TAI98 2895TAI98 2896TAI98 2897TAI98 2899TAI98 2901TAI98 2906TAI98 2911TAI98 2912TAI98 2913TAI98 2914TAI98 2915TAI98 2916TAI98 2917TAI98 2943TAI98 2945TAI98 2949TAI98 3254TAI98 TW574698 TW464398 TW609298
GenBank Accession No. AF119795 AF119796 AF136379 AF176044 AF286489 AF286490 AF286491 AF286492 AF286493 AF286494 AF286495 AF286496 AF286497 AF286498 AF286499 AF286500 AF286501 AF286502 AF286503 AF286504 AF286505 AF286506 AF286507 AF286508 AF286509 AF286510 AF286511 AF286512 AF286513 AF286514 AF286515 AF286516 AF286517 AF286518 AF286519 AF286520 AF286521 AF286522 AF286523 AF286524 AF286525 AF286526 AF286527 AF286528 AF286529 AF286530 AF286531 AF304457 AF304458 AF304459
entrez/query.fcgi) during 1997 to 1998 are selected for this research, as listed in Table 2. Among 50 nucleotide sequences, 48 sequences are from Taiwan and 2 from Singapore (1423SIN98 and 1424SIN97). Most of these datasets (49 sequences) are in 1998 owing to the outbreak of this epidemic disease in the year 1998 in Taiwan, and one sequence from 1997 (1424SIN97). Among these sequences, seven of them are
Description Complete CDS (CGU,TW) Complete CDS (CGU,TW) Complete CDS (CKUH,TW) Complete CDS (CDC,TW) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Partial CDS (CDC,USA) Complete CDS (CKUH,TW) Complete CDS (CKUH,TW) Complete CDS (CKUH,TW)
complete CDS, including four sequences from National Cheng Kung University Hospital (NCKU98, TW574698, TW464398, and TW609298),24,25 two from Chang Gung University (TW227298 and TW208698),21 and one from CDC (Taipei, Taiwan), and others are all partial CDS (VP1). For more details about these sequences, please consult the NCBI website and the associate references.
January 8, 2008 17:13 00039
298
T.-C. Shih & P.-Y. Chen
Likelihood Analysis of Recombination in DNA Maximum-likelihood method, first established by Felsenstein,9 is fundamental and powerful in phylogeny
P (t) =
A
(It’s called the F81 model). The work also describes a novel Markov model for DNA substitution, which is an extension of the Jukes and Cantor model (JC model). The substitution probability matrix P (t) is shown as following equation:
C
G
T
πA f (t) + 1 − f (t)
πC f (t)
πG f (t)
πT f (t)
πA f (t)
πC f (t) + 1 − f (t)
πG f (t)
πT f (t)
πA f (t)
πC f (t)
πG f (t) + 1 − f (t)
πT f (t)
πA f (t)
πC f (t)
πG f (t)
πT f (t) + 1 − f (t)
(1)
In Eq. 1, each item represents a substitution probability. For instance, row 3:column 2 implies the probability to change from G to C in time t, where f (t) is a function of the evolution time t. After theoretical calculations, though tedious but straight forward, one can get the conclusion as following equations: P (t) =
A
C
G
T
πA (1 − e−αt ) + e−αt
πC (1 − e−αt )
πG (1 − e−αt )
πT (1 − e−αt )
πA (1 − e−αt )
πC (1 − e−αt ) + e−αt
πG (1 − e−αt )
πT (1 − e−αt )
πA (1 − e−αt )
πC (1 − e−αt )
πC (1 − e−αt ) + e−αt
πT (1 − e−αt )
πA (1 − e−αt )
πC (1 − e−αt )
πG (1 − e−αt )
πC (1 − e−αt ) + e−αt
(2)
In brief, this matrix can be written as: Pij (t) = e
−αt
δij + (1 − e
−αt
)πj
(3a)
seen in Eqs. 2 and 3) over sites follow the Gamma distribution26 : f (r) = β α Γ(α)−1 e−βr rα−1 ,
or Pij (t) = πj + e−αt (δij − πj )
(3b)
Here, the delta is Kronecker’s delta, as Eq. 4 δij =
1 if i = j 0 if i = j
(4)
and the overall substitution rate is α as well. For more details, please consult Dr J. Felsenstein’s famous paper in Ref. 9. While the assumption in the JC model is that probability of change from any state to and different state is always equal, in the F81 model, it is assumed that the probability of change from any state i to state j is proportional to the frequency of state j. However, this method assumes that the rate of substitution is the same at different nucleotide sites and is unrealistic. In this study, we introduce likelihood analysis of recombination in DNA, under the assumption that substitution rates r (while the overall rate is α as the F81 model, which have discussed above, can be
r > 0.
(5)
The mean of substitution rates r is E(r) = α/β as well as the variance is Var(r) = α/β 2 . β is a trivial scale factor, and in order to avoid the use of too many parameters, it’s restricted the mean of the distribution to be 1 and set α = β. Thus, the Gamma distribution is related to a single parameter α, which determines the extent of rate variation. A small α suggests that rates differ significantly over sites, while a very large α means roughly equal rates. Other assumptions, such as the independence of nucleotide substitutions at different sites and possible variation of substitution rates along different lineages, are the same as Felsenstein.9 In this research, the HKY model10 is induced. The HKY model allows for a different rate of transitions and transversions as unequal frequencies of the four nucleotides (base frequencies). Furthermore, the transition to transversion ratio (Ts/Tv) is needed in this model to calculate the substitution. Before this, some researchers have already build models to evaluate the substitution, such as Jukes-Cantor model (JC69), Kimura model (K2P), and F81 model (F81). If the Ts/Tv is set to 0.5 as well, then it becomes equivalent
January 8, 2008 17:13 00039
The Recombination of Human Enterovirus 71
Phylogeny Analysis for Recombination in EV71 Virus-The Neighbor-Joining Trees In this section, neighbor-joining trees19 are selected because it is precise and well developed. This method, briefly speaking, is to find pairs of operational taxonomic units (neighbors) that minimize the total branch length at each stage of clustering of operational taxonomic units starting with a starlike tree can quickly be obtained by using this method. Though this method is quite time consuming, and when the number of operational taxonomic units (OTUs) is large, only a few percentage of all possible topologies will be examined. However, some methods in which the process of searching for the minimum evolution tree is built into the algorithm, so that a unique final topology is obtained automatically. Furthermore, this method also does not necessarily produce the minimum-evolution tree, and will be applicable to any type of evolutionary distance data. Generally speaking, this work is to establish neighbor-joining trees from several candidate sequences selected by the prior method (Likelihood Analysis of recombination in DNA, please see the former topic). To construct the neighbor-joining trees separately from either side of the recombination point which have been predicted from the above method (likelihood analysis of recombination in DNA), and then the recombination evidence will be easily unveiled. Simplifying the issue to three representative sequences (the recombinant and two parents) will make the test of phylogenetic disparity more conservative. All the neighbor-joining trees used here were constructed with the MEGA2,13 which is a very useful and powerful tool in constructing phylogeny trees.
40
30 Likelihood Ratio
to the Jukes-Cantor model; if the base frequencies are set equal, then it becomes equivalent to the Kimura 2parameter model; if the Ts/Tv is set to 0.5 and the base frequencies are not equal, then this model is equivalent to F81 model.
299
20
10
0 600
650
700
750
800
850
900
Recombination point (b.p.)
Fig. 1 Plots of the likelihood ratio versus the recombination point prediction. Each dotted point represent one dataset of enterovirus vp1 sequences.
includes three sequences in it, the putative recombinant and two reference sequences. Figure 1 unveils the result of likelihood analysis, which shows the distribution of likelihood ratio of the EV71 vp1 concisely and clearly as well. The highest ten percent datasets (about 1000 datasets) are chosen to this figure, for that one can find out the most possible recombination points more easily. From this figure, one can figure out that the ensemble region is mainly located in the region between 670 bp ∼ 700 bp, and a smaller cluster is between 880 ∼ 890 bp. After carefully cross-validation, I select the following eight sequences: 2734TAI98 (AF286494), TW227298 (AF119795), 1423SIN98 (AF286489), 2735ATAI98 (AF286495), 2894TAI98 (AF286514), 2901TAI98 (AF286519), 3254TAI98 (AF286531), 2850TAI98 (AF286499). From this calculation, the highest likelihood ratio scores is: (2734TAI98, TW227298, 1423SIN98). However, the identification of the recombinant and its parents is difficult just by using likelihood. In order to examine the phenomena of recombinant, it is necessary to construct the associate phylogeny trees for identifying the recombinant and its parents. In the next section, these eight sequences together will be established phylogeny trees for finding support for recombination evidence.
RESULTS AND DISCUSSION The Likelihood Analysis
The Neighbor-Joining Trees Analysis
The likelihood analysis provides the apparent evidence for the recombinant among EV71 viruses. In order to determine which of these sequences were the most likely recombinants, we calculate all possible collections to screen the candidates. In this research, there are 7350 datasets (C350 = 7350) to be calculated. Each dataset
The neighbor-joining trees analysis has long been considered as the easy and apparent way to examine the recombinant phenomena. To construct maximum likelihood trees for each of the two proposed recombinant regions, choosing the nucleotide 690 as the breakpoint, along with a bootstrap analysis involving 1000
January 8, 2008 17:13 00039
300
T.-C. Shih & P.-Y. Chen
(A) Neighbor-joining tree (1 ∼ 690 b.p.)
(B) Neighbor-joining tree (691 ∼ 891 b.p.) Fig. 2 Phylogenetic supports for recombination in EV71 virus. The neighbor-joining trees (with associated bootstrap values) reveal the evidence of recombination. The nucleotide 690 as the breakpoint was identified in prior maximum likelihood analysis.
replicate neighbor-joining trees. Figure 2(a) unveils the maximum likelihood tree, which is constructed from eight candidates that are parts of VP1 sequences (from 1–690 bp). In this figure, TW227298 and 1423SIN98 are beneath together, however 2734TAI98 is far from them. It means TW227298 and 1423SIN98 are much similar than 2734TAI98. Then, in Fig. 2(b), we also construct the neighbor-joining tree from eight candidates that are other parts of VP1 sequences (from 691–891 bp) in the same way. In this figure, TW227298 and 2734TAI98 are close to each other, and 1423SIN98 is far away from the other two sequences. That implies TW227298 and 2734TAI98 are much similar than 1423SIN98. After this cross-validation, one can demonstrate that TW227298 is the recombinant and 2734TAI98 and 1423SIN98 are parents.
ACKNOWLEDGMENTS We thank our parents, all of my friends and colleagues for their valuable suggestions to improve this paper.
REFERENCES 1. Andino R, Rieckhof GE, Achacoso PL et al., Poliovirus RNA synthesis utilizes an RNA complex formed around the 5 -end of viral RNA, EMBO J 12:3587–3598, 1993.
2. Archard LC, Bowles NE, Behan PO et al., Postviral fatigue syndrome: Persistence of enterovirus RNA in muscle and elevated creatine kinase, J R Soc Med 81:326–329, 1988. 3. Archard LC, Freeke C, Richardson P et al., Persistence of enterovirus RNA in dilated cardiomyopathy: A progression from myocarditis, in Schultheiss H-P (ed.), New Concepts in Viral Heart Disease, Springer-Verlag KG, Berlin, Germany, pp. 349–362, 1998. 4. Banatvala JE, Insulin-dependent (juvenile-onset, type 1) diabetes mellitus: Coxsackie B viruses revisited, Prog Med Virol 34:33–54, 1987. 5. Barrett-Connor E, Is insulin dependent diabetes mellitus caused by Coxackie B infection? A review of the epidemiologic evidence, Rev Infect Dis 7:207–215, 1985. 6. Bowles NE, Richardson PJ, Olsen EGJ et al., Detection of coxsackie-B-virus-specific RNA sequences in myocardial biopsy samples from patients with myocarditis and dilated cardiomyopathy, Lancet 1120–1123, 1986. 7. Cambridge G, MacArthur CGC, Waterson AP et al., Antibodies to coxsackie B viruses in primary congestive cardiomyopathy, Br Heart J 41:692–696, 1979. 8. Clements GB, McGarry F, Nairn C et al., The detection of enterovirus specific RNA in serum: The relationship to chronic fatigue, J Med Virol 45:151–166, 1995. 9. Felsenstein J, Evolutionary trees from DNA sequences: A maximum likelihood approach, J Mol Evol 17:368– 376, 1981. 10. Hasegawa M, Kisino H, Yano T, Dating the human human-ape splitting by a molecular clock of mitochondrial DNA, J Mol Evol 22:160–174, 1985. 11. Ho M, Chen ER, Hsu KH et al., An epidemic of enterovirus 71 infection in Taiwan. Taiwan enterovirus epidemic working group, N Engl J Med 341:929–935, 1999. 12. Kitamura N, Semler BL, Rothberg PG et al., Primary structure, gene organization and polypeptide expression of poliovirus RNA, Nature 291:547–553, 1981. 13. Kumar S, Koichiro T, Ingrid BJ et al., MEGA2: Molecular evolutionary genetics analysis software, Bioinformatics 17:1244–1245, 2001. 14. Li CC, Yang MY, Chen RF et al., Clinical manifestations and laboratory assessment in an enterovirus 71 outbreak in southern Taiwan, Scand J Infect Dis 34:104–109, 2002. 15. Liu CC, Tseng HW, Wang SM et al., An outbreak of enterovirus 71 nfection in Taiwan, 1998: Epidemiologic and clinical manifestations, J Clin Virol 17:23–30, 2000. 16. Melnick JL, Tagaya I, Magnus H, Enteroviruses 69, 70, and 71, Intervirology 4:369–370, 1974. 17. Melnick JL, Enteroviruses: Polioviruses, coxackieviruses, echoviruses, and newer enteroviruses, in Fields BN, Knipe DM, Howley PM et al., Fileds Virology, 3rd edn. Lippinocoff-Raven Publishers, Philadelphia, pp. 655–712 (1996). 18. Rossman MG, Arnold A, Erickson EA et al., Structure of a human common cold virus and functional relationship to other piconaviruses, Nature 317:145–153, 1985. 19. Saitou N, Nei M, The neighbor-joining method: A new method for reconstructing phylogenetic trees, Mol Biol Evol 4:406–425, 1987.
January 8, 2008 17:13 00039
The Recombination of Human Enterovirus 71 20. Sharief MK, Hentges R, Ciardi M, Intrathecal immune response in patients with the post-polio syndrome, N Engl J Med 325:749–755, 1991. 21. Shih SR, Ho MS, Lin KH et al., Genetic analysis of enterovirus 71 isolated from fatal and non-fatal cases of hand, foot and mouth disease during an epidemic in Taiwan, Virus Res 68:127–136, 2000. 22. Toyoda H, Nickliu MJH, Murray MG et al., A second virus-encoded proteinase involved in proteolytic processing of poliovirus polyprotein, Cell 45:761–770, 1986. 23. Woodall CJ, Riding MH, Graham DI et al., Sequences specific for enteroviruses detected in spinal cord from patients with motor neurone disease, Br Med J 308: 1541–1543, 1994.
301
24. Yan JJ, Wang JR, Liu CC et al., An outbreak of enterovirus 71 infection in Taiwan 1998: A comprehensive pathological, virological, and molecular study on a case of fulminant encephalitis, J Clin Virol 17:13–22, 2000. 25. Yan JJ, Su IJ, Chen PF et al., Complete genome analysis of enterovirus 71 isolated from an outbreak in Taiwan and rapid identification of enterovirus 71 and coxackievirus A16 by RT-PCR, J Med Virol 65:331– 339, 2001. 26. Yang Z, Maximum–likelihood estimation of phylogeny from DNA sequences when substitution rates differ over sites, Mol Biol Evol 10:1396–1401, 1993.
