A universal PCR primer to detect members of the Potyviridae and its use to examine the taxonomic status of several members of the family

Summary.  A universal primer (Sprimer: 5^′-GGX AAY AAY AGY GGX CAZ CC-3^′, X = A, G, C or T; Y = T or C; Z = A or G), designed from the consensus sequences that code for the conserved sequence GNNSGQP in the NIb region of members of the family Potyviridae, was used to amplify by RT-PCR the 3′-terminal genome regions from infected plant samples representing 21 different viruses in the family. Sequencing of some of the fragments (c. 1.7 kb) showed that the type strain (ATTC PV-107) of Oat necrotic mottle virus is not a distinct species in the genus Rymovirus, but is synonymous with Brome streak mosaic virus (genus Tritimovirus) and that Celery mosaic virus is a distinct member of the genus Potyvirus not closely related to any other sequenced species. Potyviruses infecting crops in China were also investigated, showing that viruses on cowpea and maize in Hangzhou, Zhejiang province were respectively Bean common mosaic virus and Sugarcane mosaic virus and that one on garlic in Nanjing, Jiangsu province was Onion yellow dwarf virus. Fragments were also sequenced from Chinese isolates of Lettuce mosaic virus and Soybean mosaic virus (from Hangzhou), Turnip mosaic virus (2 different isolates from Zhejiang province) and RNA1 of Wheat yellow mosaic virus (from Rongcheng, Shandong province). Received June 30, 2000 Accepted September 28, 2000     

The genome organization of the broad bean necrosis virus (BBNV)

Summary.  The genome of the broad bean necrosis virus Oita-isolate (BBNV-O) [RNA1 (6.0 kb), RNA2 (2.8 kb) and RNA3 (2.4 kb)] was cloned and sequenced. Computer analysis indicates that methyltransferase, helicase and RNA-dependent RNA polymerase (RdRp) motifs are present in RNA1. The viral capsid protein (CP) cistron is located at the 5′ terminal end of RNA2 and the Mr of CP (20 K) is close to that determined by SDS-PAGE analysis. An ochre codon (UAA) in the CP cistron is thought to be partially suppressed to produce a large readthrough protein. RNA3 possesses typical motifs of triple gene block proteins, which are also reported in several other plant viruses. The furovirus genome organization and phylogenetic analysis using RdRp and CP amino acid sequences suggest that BBNV is closely related to potato mop-top virus (PMTV), but is relatively distantly related to other furoviruses. The data also suggest that the genus Furovirus should be separated into several genera: the prototypical genus Furovirus, which excludes the following viruses: the PMTV group including BBNV; the beet necro- tic yellow vein virus (BNYVV) group; and the peanut clump virus (PCV) group. Received November 14, 1997 Accepted Febuary 12,1998     

Biological,molecular, and serological studies of a novel strain of grapevine leafroll-associated virus 2

In California, a novel closterovirus was detected in “Redglobe” grapevine, associated with graft incompatibility and given a trivial name “Grapevine rootstock stem lesion associated virus (GRSLaV).” The biological properties of the putative virus were ascertained when asymptomatic yet infected Redglobe scion buds were graft-inoculated onto test plants of Cabernet Sauvignon propagated on 18 different rootstocks. It proved lethal on test plants growing on rootstocks 1616C, 5BB, 5C, 3309C, and 1103 P, whereas latent infections occurred on the remaining scion-rootstock combinations. In contrast, GLRaV-2 type (type strain) produced only typical leafroll symptoms. In a different experiment, GLRaV-2 type was successfully sap-transmitted to N. benthamiana, whereas sap transmission of GRSLaV was unsuccessful. Double-stranded RNA was extracted from infected Redglobe grapevines, cloned, sequenced, and determined a genome length of 16,527 nucleotides. Computer-assisted analysis of open-reading frames (ORFs) revealed a genome organization typical of monopartite viruses in the genus Closterovirus with nine ORFs (range 71–79% identity) with GLRaV-2 type, the closest similar virus species within the family Closteroviridae. Also the 3′-UTR of GRSLaV consisted of 223 nucleotides with an extended oligo(A) tract similar to that of GLRaV-2 type, Beet yellow stunt virus, and Beet yellows virus. Recombinant GRSLaV coat protein was expressed in E. coli, purified, and immunized a rabbit to produce polyclonal antiserum. Serological data matched the molecular data, whereby exposed plant tissue extracts of grapevines infected by both viruses (GRSLaV and GLRaV-2) reacted positively with homologous and heterologous viral antisera but not with healthy grapevine extracts in ELISA and Western blot tests. Based on the comparative sequence data and shared antigens, GRSLaV is now considered a strain of GLRaV-2 and redesignated as Grapevine leafroll associated virus-2 Redglobe (GLRaV-2RG). Primers specific for GLRaV-2RG were developed, which did not amplify GLRaV-2 type strain. When both sets of specific primers were used in assays of different grapevine collections, the incidence of the respective viruses varied considerably, e.g., 1.7 and 13.5%, respectively, for GLRaV-2RG and GLRaV-2 type.     

Genome structure and complete sequence of genomic RNA of Daphne virus S

Summary. The complete genomic nucleotide sequence and structure of Daphne virus S (DVS), a daphne-infecting member of the genus Carlavirus, were determined. The genome of DVS was 8,739 nucleotides long, excluding the poly (A) tails. The genome of DVS contained six open reading frames coding for proteins of Mr 227 kDa (viral replicase), 25 kDa, 11 kDa and 7 kDa (triple gene block TGB) proteins 1, 2 and 3), 35 kDa (coat protein; CP), and 12 kDa from the 5′ to 3′ ends; respectively. This is the typical genome structure of members of the genus Carlavirus. Overall amino acid sequence similarities for the six ORFs of DVS were from 58.5% to 13.2% to those of the other carlaviruses. The 227 kDa replicase of DVS shared 45.5–39.2% amino acid similarities to that of 8 other known carlaviruses. Results from phylogenetic analyses of viral replicases and CPs demonstrated that DVS is a close relative of Helenium virus S and Chrysanthemum virus B. A total of 13 isolates of DVS shared 100–95.9% identities for the amino acid level and 99.5–81.0% identities for the nucleotide level. This is the first report of the complete genome sequence and structure of DVS and supports the conclusion that DVS is a typical species of the genus Carlavirus.     

Characterisation of a potyvirus and a potexvirus from Chinese scallion

Summary.  Molecular analyses of viruses infecting Chinese scallion (Allium chinense G. Don) showed that the plants did not contain any of the poty-, carla- or allexiviruses that are common in garlic plants in China. The complete sequences of a potyvirus and a potexvirus were determined and these were shown to represent different viruses from any in the databases. They could be transmitted mechanically to scallion but not to other Allium species (including garlic) or to Narcissus. The potyvirus, tentatively named Scallion mosaic virus, has a distant relationship (c. 62% nucleotide identity over the entire genome) to Turnip mosaic virus and Japanese yam mosaic virus, with which it grouped in phylogenetic analyses. Its genome is 9324 nts long, encoding a 341.3 kDa polyprotein of 3001 amino acids. The potexvirus, tentatively named Scallion virus X, has a genome 6987 nts long and its organisation was similar to that of the other potexviruses but with only 46.3–63.2% nucleotides identical to them. It is most closely related to Narcissus mosaic virus but phylogenetic analyses indicate that it should be considered a distinct species. Neither of the viruses have been detected in garlic, although the two host plants are closely related. Received September 28, 2001 Accepted December 20, 2001     

A new small temperate DNA phage BcP15 isolated from <Emphasis Type="Italic">Burkholderia cepacia</Emphasis> DR11

Summary. A Burkholderia cepacia DR11 strain was isolated during the survey of microorganisms from coastal water of deltaic Sunderbans. This strain always released temperate phage BcP15 into culture supernatant. UV irradiation of the strain also induced phage induction. The phage titer was 2.3 × 10⁸. New temperate phage BcP15 has unusual structure. It has a hexagonal head, 65 nm in diameter and a tail 200 nm long, attached with single thick wavy tail fiber (424–705 nm). Phage DNA is double stranded 11.9 kb long. Southern hybridization result indicated that the phage DNA was in lysogenic state into the B. cepacia DR11 genome. SDS-PAGE of phage protein showed two major bands of molecular weight 20 kDa and 40 kDa.     

Identification and characterization of a novel tospovirus species using a new RT-PCR approach

Summary.  A novel tospovirus serologically distinct from all established tospo- virus species was found in Thailand in Physalis minima L. The S RNA of this virus was cloned by a new RT-PCR approach revealing a nucleotide sequence of 3257 nucleotides. The ambisense RNA segment encoded a nonstructural protein (NSs) of 469 amino acids, with a predicted Mr of 53.2 kDa, and a nucleoprotein (N) of 279 amino acids and a Mr of 31.0 kDa, so far the largest N protein known for any tospovirus species. N protein sequence comparisons revealed closest relationship to the species Watermelon bud necrosis virus (58% identity), Watermelon silver mottle virus and Peanut bud necrosis virus (57%) and a distant relationship to Peanut yellow spot virus (23%) and Peanut chlorotic fanspot virus (22%). Received March 27, 2000 Accepted July 27, 2000     

Hibiscus virus S is a new subgroup II tobamovirus: evidence from its unique coat protein and movement protein sequences

Summary.  The coat protein (CP) and movement protein (MP) sequences of a new tobamovirus infecting Hibiscus rosa-sinensis L were determined. The CP gene encodes 163 amino acid (aa) residues and with a theoretical molecular weight of 18.19 kDa. The MP gene encodes 282 amino acids and its theoretical molecular weight is 30.36 kDa. The nucleotide (nt) and aa sequences of the CP were 46.88 % to 51.63 % and 45.34 % to 57.06 % identical to other tobamoviruses, respectively. The nt and aa sequence identities of MP ranged from 38.81 % to 43.90 % and 30.85 % to 37.88 %, respectively. The predicted virion origin of assembly (OAS) was located in the CP gene. Phylogenetic trees generated based on the nt and aa sequences of both CP and MP genes indicate that this new virus clusters with members of subgroup II of tobamoviruses. Although this hibiscus virus shared a high nt and aa sequence identity with Sunn-hemp mosaic virus (SHMV), Western analysis showed that it is serologically unrelated to SHMV. We propose the name Hibiscus virus S (HVS) for this Singapore isolate. This is the first report on partial nt sequence of a tobamovirus that infects hibiscus. Received November 22, 2001; accepted February 28, 2002 Published online June 21, 2002     

Molecular analysis of the genome segments S1, S4, S6, S7 and S12 of a Rice gall dwarf virus isolate from Thailand; completion of the genomic sequence

Summary The complete nucleotide sequences of the double-stranded RNA segments S1, S4, S6, S7 and S12 of the genome of a Rice gall dwarf virus (RGDV) isolate from Thailand were determined. The segments consisted of 4505, 2622, 1648, 1652 and 853 nucleotides, encoding putative proteins of 1458, 725, 489, 511 and 206 amino acids with molecular masses of approximately 166, 80, 53, 59 and 24 kDa, respectively. Homology searches indicated that each of the putative proteins has a counterpart in isolates of Rice dwarf virus (RDV) and Wound tumor virus, two other species in the genus Phytoreovirus. However, no similarities were found to other registered sequences, including those of other viruses that belong to the family Reoviridae. The identities between homologous structural proteins of RGDV and RDV ranged from 34 to 51% and were thus higher than those between homologous non-structural proteins of RGDV and RDV (16–37%). Among the nonstructural proteins, the highest amino acid sequence identity (37%) was observed for RGDV Pns11 and RDV Pns10, a constituent of tubular inclusions. This observation suggests that a specific amino acid backbone might be required for maintaining not only the three-dimensional structure of virions but also that of inclusions. The entire sequence of the RGDV genome is now available.     

The nucleotide sequence and genome organization of Japanese iris necrotic ring virus, a new species in the genus Carmovirus

Summary.  The genome of Japanese iris necrotic ring virus (JINRV) consists of a positive-sense ssRNA of 4014 nucleotides with six major open reading frames (ORFs). A 5′-non-coding region of 31 nucleotides precedes the first initiation codon. Like Carnation mottle virus (CarMV), the 5′-proximal three ORFs encode a 26 kDa protein (p26) and two readthrough proteins, i.e. an 85 kDa putative RNA replicase (p85) and a 99 kDa protein (p99). The central ORF encodes a small 8 kDa protein (p8). The 3′-proximal ORF encodes a 38 kDa capsid protein (p38). Another ORF encoding a 12 kDa protein (p12) overlaps the p99 ORF. JINRV RNA treated with bacterial alkaline phosphatase and tobacco acid pyrophosphatase could not be ligated to an oligoribonucleotide using T4 RNA ligase, indicating that the 5′ end of the viral RNA is uncapped. The 3′ end is not polyadenylated. Comparison of the genomic organization and the predicted amino acid sequences with those of other viruses confirmed that JINRV should be classified as a member of the genus Carmovirus, family Tombusviridae. Accepted September 23, 1999     

Indian citrus ringspot virus: a proposed new species with some affinities to potex-, carla-, fovea- and allexiviruses

Summary.  An isolate of Indian citrus ringspot virus from Kinnow mandarin in northern India had flexuous particles with evident cross-banding and a modal length of 650 nm. It was mechanically transmitted to five herbaceous hosts including Phaseolus vulgaris cv Saxa, in which it became systemic. In thin sections, virus particles were observed in the cytoplasm of parenchyma cells but no specific inclusions were seen. The virus was purified from infected Saxa bean leaves and an antiserum prepared. There was no serological cross-reaction with representative allexi-, capillo-, potex- and trichoviruses, except a faint one-way reaction with Potato virus X. Purified virus yielded a major band, the presumed coat protein (CP), of about 34 kDa, and a single ssRNA of about 7.5 kb, which was infectious. Two ORFs encoding putative proteins of 34 kDa and 23 kDa were located in the 3′ part of the RNA. The product of the 34 kDa ORF was confirmed as the CP by expression in E. coli. The derived amino acid sequence of the CP contained some short motifs similar to those of potex-, fovea-, carla- and allexiviruses but otherwise there was no strong similarity to any of these. The 23 kDa ORF contained a zinc finger-like sequence, as in similar ORFs in carla- and allexiviruses but overall amino acid homology with these was low. The virus does not appear to fall into any known genus. A new species is proposed. Serological and molecular diagnostic reagents were prepared. Received July 14, 1999/Accepted February 10, 2000     

Sequence analysis demonstrates that <Emphasis Type="Italic">Onion yellow dwarf virus</Emphasis> isolates from China contain a P3 region much larger than other potyviruses

Summary.  The complete sequence of an isolate of Onion yellow dwarf virus (OYDV) from Yuhang, Zhejiang province, China, was determined. It was 10538 nts in length and was predicted to encode a polyprotein 3403 amino acids (aa) long with a calculated M_r of 385.1 kDa. The predicted P3 protein (530 aa) was larger than that of any of the potyviruses sequenced to date (344–378 aa). The additional sequence occurs at the N-terminus of the protein, does not represent a duplication from elsewhere in the OYDV genome and could not be matched to any other sequences in the databases. Similar sequences were found in 4 other Chinese OYDV isolates. Phylogenetic analysis of the amino acid sequences of the polyprotein showed that OYDV is distantly related to Pea seed-borne mosaic virus and the potyviruses of grasses and cereals. Received November 26, 2002; accepted February 4, 2003 Published online April 2, 2003     

Detection and characterisation of a second potyvirus from Thunberg fritillary in China

Summary. Plants of Thunberg fritillary (Fritillaria thunbergii Miq.) from Zhejiang Province, were found to be co-infected with two distinct potyviruses. One was an isolate of the recently reported Thunberg fritillary mosaic virus (TFMV; Wei et al., (2005) Arch Virol 150: 1271–1280), while the other was a distinct virus that did not react with TFMV antiserum nor with antisera to 17 other potyviruses, except for a weak reaction with antibodies produced to soybean mosaic virus (SMV) Pinellia strain. Both viruses could be transmitted mechanically to their original host but not to any of a range of commonly used indicator plants. No local lesion host was identified that would enable the viruses to be propagated independently. The complete sequences of both viruses were determined; that of the new virus (9656 nt) had the typical genome organisation and recognised sequence motifs of a potyvirus, encoding a putative polyprotein of 351 kDa. Phylogenetic analysis, sequence comparisons, and the pattern of polyprotein cleavage sites all indicated that it was a member of the Bean common mosaic virus subgroup. The most closely related species are Soybean mosaic virus and Wisteria vein mosaic virus, with 68–69% amino acid identity between their polyproteins. This is sufficiently different for the new virus to be regarded as a distinct species, which we have tentatively named Fritillary virus Y.     

Genetic organisation of <Emphasis Type="Italic">Iris yellow spot virus</Emphasis> M RNA: indications for functional homology between the G<Subscript>(C)</Subscript> glycoproteins of tospoviruses and animal-infecting bunyaviruses

Summary.  The complete nucleotide sequence (4838 nucleotides) of Iris yellow spot virus (IYSV) M RNA indicates, typical for tospoviruses, the presence of two genes in ambisense arrangement. The vRNA ORF codes for the potential cell-to-cell movement (NSm) protein (34.8 kDa) and the vcRNA ORF for the viral glycoprotein (G1/G2) precursor (128.6 kDa). Multiple sequence alignment of the NSm and G1/G2 precursor proteins of IYSV with those of other tospoviruses, showed highest homologies to Peanut bud necrosis virus (PBNV) and Watermelon silver mottle virus (WSMV). The potential cell-to-cell movement protein of tospoviruses is highly conserved (40–70% identity), with the exception of the first 60 N terminal amino acids, a domain that clearly diverged. For the G1 and G2 viral glycoproteins, blast searches revealed a significant homology between the C-terminally located tospoviral G1 (G_(C)) protein with the counterpart of the animal-infecting bunyaviruses, suggesting a functional homology for these proteins. Received January 15, 2002; accepted July 10, 2002     

Apricot latent virus: a new species in the genus Foveavirus

Summary.  Extraction of viral double-stranded RNA from peach leaves infected with Apricot latent virus (ALV) followed by molecular cloning of synthesized cDNA and its sequencing, suggested that ALV is a new virus, whose coat protein (CP) coding region contains Apple stem pitting virus (ASPV)-related sequences. The sequenced portion of the ALV genome (1444 nt) includes the putative CP gene and the 3′ non-translated region. The 5′ portion of this fragment (1-651 nt) is highly distinct whereas the 3′ portion is 77% identical to the corresponding region of ASPV. Molecular hybridization experiments using a cRNA probe to ASPV with ALV-infected leaf tissue extracts also revealed that the genome of ALV contains nucleotide sequences related to that of ASPV. Western blots of tissue extracts indicated that ALV coat protein reacted with polyclonal antiserum against ASPV; however, the ALV CP differs in size from that of ASPV. ALV was graft-transmitted to several Prunus rootstocks. Based on the available sequence data, serological observations and bioassays we propose that ALV is a new species in the genus Foveavirus, typified by ASPV. ALV-specific PCR-primers and viral-specific cRNA probes developed in this investigation may be useful for detecting the virus and for studying its epidemiology and geographical distribution. Received February 10, 2000/Accepted April 3, 2000     

Complete nucleotide sequence and experimental host range of <Emphasis Type="Italic">Okra mosaic virus</Emphasis>

Okra mosaic virus (OkMV) is a tymovirus infecting members of the family Malvaceae. Early infections in okra (Abelmoschus esculentus) lead to yield losses of 12–19.5%. Besides intensive biological characterizations of OkMV only minor molecular data were available. Therefore, we determined the complete nucleotide sequence of a Nigerian isolate of OkMV. The complete genomic RNA (gRNA) comprises 6,223 nt and its genome organization showed three major ORFs coding for a putative movement protein (MP) of M_r 73.1 kDa, a large replication-associated protein (RP) of M_r 202.4 kDa and a coat protein (CP) of M_r 19.6 kDa. Prediction of secondary RNA structures showed three hairpin structures with internal loops in the 5′-untranslated region (UTR) and a 3′-terminal tRNA-like structure (TLS) which comprises the anticodon for valine, typical for a member of the genus Tymovirus. Phylogenetic comparisons based on the RP, MP and CP amino acid sequences showed the close relationship of OkMV not only to other completely sequenced tymoviruses like Kennedya yellow mosaic virus (KYMV), Turnip yellow mosaic virus (TYMV) and Erysimum latent virus (ErLV), but also to Calopogonium yellow vein virus (CalYVV), Clitoria yellow vein virus (CYVV) and Desmodium yellow mottle virus (DYMoV). This is the first report of a complete OkMV genome sequence from one of the various OkMV isolates originating from West Africa described so far. Additionally, the experimental host range of OkMV including several Nicotiana species was determined. The nucleotide sequence data reported in this article have been submitted to the Genbank nucleotide sequence database and have been assigned the accession number EF554577.     

Complete sequence of Fig fleck-associated virus, a novel member of the family Tymoviridae

The complete nucleotide sequence and the genome organization were determined of a novel virus, tentatively named Fig fleck-associated virus (FFkaV). The viral genome is a positive-sense, single-stranded RNA 7046 nucleotides in size excluding the 3′-terminal poly(A) tract, and comprising two open reading frames. ORF1 encodes a polypeptide of 2161 amino acids (p240), which contains the signatures of replication-associated proteins and the coat protein cistron (p24) at its 3′ end. ORF2 codes for a 461 amino acid protein (p50) identified as a putative movement proteins (MP). In phylogenetic trees constructed with sequences of the putative polymerase and CP proteins FFkaV consistently groups with members of the genus Maculavirus, family Tymoviridae. However, the genome organization diverges from that of the two completely sequenced maculaviruses, Grapevine fleck virus (GFkV) and Bombix mori Macula-like virus (BmMLV), as it exhibits a structure resembling that of Maize rayado fino virus (MRFV), the type species of the genus Marafivirus and of Olive latent virus 3 (OLV-3), an unclassified virus in the family Tymoviridae. FFkaV was found in field-grown figs from six Mediterranean countries with an incidence ranging from 15% to 25%.     

Characterisation of several heterogeneous species of defective RNAs derived from RNA 3 of cucumber mosaic virus

Summary. Preparations of double-stranded RNAs (dsRNAs) extracted from Nicotiana tabacum cv Xanthi plants infected with a subgroup IB isolate of Cucumber mosaic virus (CMV) were found to contain a heterogeneous population of defective RNAs (D-RNAs) derived from RNA 3. Characterised D-RNAs ranged in size from 1.5 to 1.9 kb and were derived either by a single in-frame deletion within the 3a or 3b genes or by means of double in-frame deletions within both genes. Also, northern blot hybridisation showed two other types of RNA derived from RNA 3: (a) RNA species of ca. 0.7 kb containing the 3′-terminus but lacking the 5′-terminus, which could be 3′-coterminal subgenomic of D-RNAs derived from the 3b gene and (b) RNA species of unknown origin of ca. 0.8 kb containing the 5′-terminus but lacking the 3′-terminus.     

Characterization and in vitro translation analysis of pelargonium flower break virus

Summary.  Pelargonium flower break virus (PFBV) is one of the common viruses in the glasshouses of Western Europe and has been assigned to the genus Carmovirus. A Spanish isolate obtained from nursery-grown Pelargonium zonale plants (PFBV-m) has been characterized. The molecular weight of genomic RNA and coat protein of PFBV-m were determined to be 1.36 × 10⁶ (corresponding to approximately 4 kb) and 36,000, respectively. Only genomic-size RNA was encapsidated in PFBV virions; making necessary to purify double-stranded RNA from infected tissue in order to detect putative PFBV subgenomic RNAs. PFBV RNA directed the synthesis of a major polypeptide of 34 kDa and three other relevant polypeptides of estimated sizes 88–90 kDa, 42 kDa and 35–36 kDa. Antisera specific to PFBV immunoprecipitated the in vitro translated 35–36 kDa polypeptide indicating that this polypeptide is the PFBV coat protein. The PFBV in vitro translation pattern was very similar to that of CarMV although the relative levels of translated coat protein differed dramatically between the two viruses, most probably due to the lack of encapsidation of subgenomic PFBV. In vitro translation studies with a different biological clone obtained from the same PFBV-m isolate revealed a prominent additional polypeptide which is postulated to be a truncation of the 5′ proximal ORF. Received November 27, 1998 Accepted March 24, 1999     

Complete nucleotide sequence of RNA-2 of Olive latent ringspot virus

Summary.  The complete nucleotide sequence of Olive latent ringspot virus (OLRSV) RNA-2 was determined. This RNA is 3969 nucleotides in length and contains a single open reading frame of 3448 nt, that encodes a polypeptide of 1146 amino acids, with a calculated M_r of 126,044. OLRSV RNA-2 has a structural organization typical of nepoviruses, with the coat protein (CP) cistron located in the C-terminal and the putative movement protein (MP) in the N-terminal regions of the polyprotein. Computer-assisted comparison of coat proteins of OLRSV and other nepoviruses disclosed relationships that tally with subgrouping based onphysicochemical properties. Received June 5, 2000 Accepted July 18, 2000