- Email: [email protected]
Sequence and analysis of the BamHI “D” fragment of Shope fibroma virus: comparison with similar regions of related poxviruses
yincs Research. 25 (1992) 117-132 @ 1992 Elsevier Science Publishers B.V. All rights reserved 0168-1702/92/$05.00
117
VIRUS 00814
Sequence and analysis of the BamHI “D” fragment of Shope fibroma virus: comparison with similar regions of related poxviruses David S. Strayer and Henry H. Jerng Department of Padlology and Cell Biology, Thomas Jefferson Utkersity Medical College, Philadelphia, PA, USA (Received 2 April 1992; revision received 26 May 1992: accepted 28 May 1992)
Summary Differences observed in the virulence of two related Ieporipoxviruses are closely tied to a particular region of their genomes. For the virulent poxvirus of this pair, malignant rabbit fibroma virus (MV), this region is the BanaH “C” fragment. which is 10.7 kb. For the avirulent poxvirus, Shope fibroma virus, SFV, this region is the corresponding BarnHI “D” fragment, which is 13.1 kb. As part of our attempt to understand the virulence of these two viruses, we sequenced these two DNA fragments. The sequence for the BarnHI “C” fragment of MV is reported elsewhere (Strayer et al., 1991). We report here the sequence for SFV’s BanzHI “D” fragment and resultant open reading frames, and compare both DNA and open reading frame structures to those of MV and other known poxviruses. The BamHI “D” fragment of SFV contains 12 open reading frames of 100 amino acids or more, arranged similarly to orf’s in MV and vaccinia. Striking similarities between SFV and MV are seen in certain parts of this restriction fragment, including substantial stretches of DNA in which the two viruses are identical. Clear homologies exist between these leporipox virus genomes and those of other related poxviruses. To understand the pathogenesis of virus infection, one must appreciate the structure of those viral genes that play important roles in infection. Malignant rabbit fibroma virus; Shope fibroma virus; Poxvirus; DNA sequence
Correspondence fcx D.S. Strayer, Deparrm~nt of Pathology and Cell Biology, Thomas Jefferson University Medical College, 1020 Locust Street, Philadelphia. PA 19107, USA. Fax: (13(215)-953-2218. Abbreviations: bp, base pairs; FPV, fowlpox virus; MV, malignant rabbit fibroma virus; MYX, rabbit myxoma virus; orf, open reading frame: SFV, Shope fibroma virus: W, vaccinia virus.
Introduction
Shope fibroma virus (SFV) was first isolated and identified in 1932 in wild rabbits in New Jersey (Shape, 1932a). At the time it was appreciated that this poxvirus was responsible for small, benign tumors that waxed and waned in rabbits over the course of several weeks. The relationship between SFV and its more virulent relative, rabbit myxoma virus (MYX), was established: prior infection with SFV protected rabbits against MYX infection (Shape, 1932b: Woodroofe and Fenner, 1965). Shortly thereafter a series of reports suggested that SFV behaved differently in very young or immunologically suppressed recipients (Ahlstrom and Andrews, 1938; Clemmensen, 1939; Duran-Reynolds, 1940; Duran-Reynolds, 1946; Bergmann et al.. 1962; Yuill and Hanson, 1964; Hurst, 1964; Allison and Friedman. 1966; Allison, 1966; Smith et al., 1973: Sell and Scott, 1981). In such animals, SFV infection caused disseminated tumors and often death. These data imply that SFV’s avirulence in normal adults is mostly a function of host immune responsiveness. Thus, differences in virulence between MYX and- SFV could be more apparent than real, SFV always producing disseminated lethal tumors in the “correct” host. We recentIy re-examined these findings. Using clonally purified stocks of SFV, we found that almost all strains of SFV caused only localized disease in steroidtreated or newborn rabbits. Only Boerlage strain SFV produced disseminated infection with necrosis in lymphoid organs in immunosuppressed and newborn rabbits. This strain, however, did not cause disseminated tumors and was not lethal (Strayer et al., 1984). Analysis of the several isolates of SFV showed that Boerlage strain SFV was genetically closer to MYX than were the other SFV strains (G. Cabirac et al., unpublished). Growing MYX and SFV together in vitro resulted in recombinant viruses of intermediate virulence (Smith, 1952). These observations indicate that SFV and MYX are sufficiently genetically related to produce recombinant viruses by co-culturing and underscore the role of viral gene structure in the course of these infections. We have been examining the virulence of malignant rabbit fibroma virus (MV). MV is a naturally occurring recombinant between SFV and MYX (Strayer et al., 19X3& the point of recombination occurred at the periphery of the virus genome. MV’s central sequences are from MYX while its terminal inverted repeat (TIR) region and a short span of unique sequences come from SFV (Upton et al., 1988). Clinically. MV is virulent, inducing disseminated tumors, severe gram-negative infection and death within 12 days. An important part of MV’s clinical behavior involves its ability to replicate in lymphocytes and suppress immune function (Strayer et al., 1987). SFV does neither of these. When MV’s 10.7 kb BamHI "c" fragment (hereafter, “C” fragment) is transferred to SFV, recombinant viruses so generated replicate in lymphocytes, suppress immune function and recapitulate most of MV’s aggressive phenotype
119
(Strayer et al., 1988, 1990). This fragment is located in the middle of the unique sequences of MV. Progressively smaller subfragments of the “C” fragment may transfer these characteristics. Two subfragments that overlap for about 700 bp are capable of transferring much of MV’s virulence to SFV: a 3.6 kb NdeI fragment and a 1.9 kb HirzcII fragment (Heard et al., 1990). The sequence of this portion of MV’s genome has been reported (Strayer et al., 1991). For Shope fibroma virus, the restriction fragment corresponding to MV’s “C” fragment is the 13.1 BamHI “D” fragment (hereafter “D” fragment). This fragment extends beyond both ends of MV’s “C” fragment. Both the C and D fragments are also homologous to the Hind111 “D” fragment of vaccinia virus (WR strain, Niles et al., 1986) and to a 10.5 kb Hind111 fragment of fowlpox virus (Tartaglia et al., 1990). Because of the importance of this region of the viral genome in the determination of virulence, we report here the sequence of SFV’s BamHI “D” fragment in its entirety and compare it to sequences of related poxviruses in this area. Small portions of this sequence have been reported elsewhere by others (Upton et al., 1991) and by ourselves (Strayer et al., 1991).
Materials
and Methods
SFVh BamHI “D” fragmerlt
The BamHI “D” fragment of SFV was obtained from cloned material generated by G. Cabirac (Cabirac, et al., 1987). A restriction map was prepared using HincII, NdeI, EcoRI, Xl101 and Hind111 (Fig. 1).
2
0
4
6
I
I
S
I
'---)z~iT~
I
7+3
4-6
9-b
10
12
I
I
I‘lkh
10 +11 + 9e
-
12
Hire
II
Em
RI
Nde
I
Hind Ill Xho I
Fig. 1. Below: restriction map of SFV BumHI “D” fragment from which sequencing was done. In this illustration. the enzymes HincII, NdeI, EcoRI. 27~01 and Hind111 are used to subdivide the 13.1 kb BarnHI fragment. Approximate sizes are shown in kb within the respective fragment boxes. Above: open reading frame organization for SFV’s BamHI “D” fragment. orf’s read from left to right are D-l. -3. -4. -5. -7, -8. -10, and -11. The others are read from right to left.
The “0” fragment was digested with the above enzymes and cloned either as singly cut or doubly cut restriction subfragments into sequencing vectors of the pGEM series (Promega). Sequencing was done by the dideoxy method (Maniatis, 1982) using first the SP6 and T7 promoters of the pGEM plasmids, plus Sequenase (USBC). (Only Sequenase was used successfully in these sequencing reactions. We tried Sequenase 11 (USBC) but were unable to obtain sequence data with it.1 Sequences further into the restriction fragments than could be obtained using the plasmid primers were sequenced using oligonucleotide primers synthesized to complement viral sequences about 50 bp upstream of the end of the previous sequence. Sequences at junctions of restriction fragments were obtained by sequencing overlapping restriction fragments. The SFV DNA sequence for this fragment has been reported to GenBank and carries accession number M74532. The sequence of MV’s BanrHI “C” fragment has accession number M32743. Sequence analysis
Sequences were assembled and analyzed using a Macintosh II computer (Apple Corp.1 with DNA Inspector IIe software (Textco, Inc.). Open reading frame (orfl recognition was performed using DNA Inspector Be. Databank
Protein and DNA sequences were compared to sequences available in GenBank via modem, using homology algorithms developed and described elsewhere (Pearson and Lipman, 1988).
Results Sequence of SFV BamHI “D” fragment
SFV’s BavrzHI “D” fragment is a 13.088 bp restriction fragment. Its sequence is shown in Fig. 2. Locations and directions of transcription for open reading frames are indicated. The organization of orf’s in SFV is very simiIar to that of other pox viruses studied, including vaccinia virus (Niles et al., 19861, malignant rabbit fibroma virus (Strayer et al, 1991) and fowlpox virus (Tartaglia et al., 19901. There are no long spaces between adjacent orfs. Often orf’s read in the same direction in different frames, or in different directions overlap each other by several bases. Open rending ~framcs
Open reading frames located in the fragment and their similarity to orf’s in MV. VV and FPV are shown in Table 1 and illustrated in Fig. 3. There are 12 orf’s in the SFV “D’” fragment. orfs D-l, -3, -4, -5, -7, -8, -10 and -11 are read left to right. The others (D-2, -6, -9 and -121 are read right to left. Except for D-12, these orf’s correspond to like-numbered orf’s in MV’s Barn “C” fragment.
Fig. 7. (Continued on pp. 122-128.) Sequence of SFV’s BumHI “D” fragment and comparison with MV’s BumHI “C” fragment. The nucleotide sequences of these two restriction fragments are shown and aligned. Positions that had to be skipped to produce alignment are shown as “.“. The larger SFV fragment is shown above the smaller MV fragment in those areas that correspond, and the numbering in plain type is indicated for the SFV sequence. Numbering for the MV sequence is shown in bold faced type. Individual lines are arranged to provide 100 SFV bases/line. Beginning and ending points for the SFV and MV open reading frames are indicated, with initial ATG’s underlined and arrows indicating the direction of transcription. The notations for SFV orf’s are above the corresponding sequences: those for MV in italics below the sequences.
Open reading frames in SFV and MV are extremely similar in size and amino acid composition. Both sets of orf’s resemble those of vaccinia virus as well as some of the fowlpox orfs that localize to the corresponding portion of fowlpox virus. The resemblance to vaccinia and fowlpox viruses is much less than that to MV. Where there is a substantial difference in orf length among the several viruses, SFV and MV both differ from vaccinia and fowlpox and resemble each other. Based on functional analyses performed for some of the open reading frames in vaccinia’s Hind111 “D” fragment, it is possible to infer the functions of SFV’s orf’s. These homologies and corresponding inferences are listed in Table 1.
As indicated above, SFV is genetically closely related to MV although their biological behavior is greatly different. As MV’s BarnHI “C” fragment is a major factor in the differential virulence of MV and SFV, we compared DNA sequences for SFV and MV in this region, covering SFV’s 13.1 kb Banz “D” fragment and MV’s 10.7 kb i?anr “C” fragment. The two sequences are very similar (Fig. 2). Locations of orf’s are identical for both viruses, save for two small orf’s present in MV but not SFV. and located near the right side of the corresponding fragment. We examined the differences between SFV and MV in this portion of the genome relative to position in the “C” fragment. At no point do the two viruses differ in more than 159%of their bases. However, substantial stretches of DNA of these two viruses are identical. This distribution is non-random (P +z 0.01). There are no base differences between SFV and MV for a stretch of 326 bp between 5301 and 5626 (SFV). This area is entirely within orf’s D-5 and C-5 for SFV and MV respectively, and also includes the CO, terminus of orfs D-6 and C-6. These latter orf’s are entirely contained within D-5 and C-5, and are transcribed in the opposite direction.
In addition, there are no differences between SFV and MV for 209 bases (8504-8712 for SFV) at the very beginning of orfs D-S and C-8. This is followed by a short stretch of 55 bp in which there are 6 differences between the two viruses resulting in two amino acid changes (I -+ V and T + I), and in turn by 1481 bp (886%10,348) without a single divergent base between the two viruses. This portion of the genome includes the carboxyl half of orf’s D-8 and C-8 (probably a subunit of DNA-dependent RNA polymerase), the entire orf’s D-9 and C-9 (probably a virion membrane protein) and almost all of orfs D-10 and C-10 (function unknown). Interestingly, none of these areas of identity between SFV and MV correspond to an area of strong homology between either of these two viruses and vaccinia. Open reading frames in this area are strongly divergent between SFV and MV on the one hand and vaccinia on the other. In those instances where orf’s for fowlpox virus have been determined and correspond to SFV, MV or VV, FPV orf’s differ from the two Ieporipox viruses more than do those of VV (Tartaglia et al., 1990).
124
Discussion
The sequence of the SFV BanrH1 “D” fragment closely resembles that of the MV BarnHI “C” fragment, is similar to vaccinia’s Hind111 “D” fragment (Niles et al., 1986) and relates somewhat more distantly to the genome of fowlpox virus as reported by Tartaglia et al. (1990). Upton and co-workers (19911, studying SFV’s capping enzyme, have sequenced the left-most portion of Kasza strain SFV. Their sequence corresponds to the first 3.5 kb of the Hi~rdIII “D” fragment, and is substantiaIly in agreement with that reported here. As shown above, the area of homology to the entire MV “c” fragment of MV is contained within SFV’s BamHI “D” fragment. The 13,088 bp of SFV’s “D” fragment is homologous to a portion of vaccinia’s 16.1 kb f?indIII “D” fragment that is completely contained within the latter restriction fragment. Open reading frame D-l from SFV corresponds to orf D-l from vaccinia with the exception that the latter orf contains 210 amino acids at the NH,-terminus not present in D-l from SFV. This difference undoubtedly represents the fact that the former orf begins near the 5’ end of the Hind111 “D” fragment of VV and that the latter orf begins near the 5’ end of SFV’s BumHI ‘“D” fragment and does not encompass the entire 5’ end of the orf.
An analogous situation obtains with SFV orf D-12, compared to W orf D-11. Both of these orf’s are read right to left. The latter orf contains 46 amino acids at the NH,-terminus that are not in SFV’s D-12. Thus, the orf from SFV probably begins 5’ to the right side of the BamHI “D” fragment of SFV and does not reflect the true size of the transcript. The organization of leporipoxviral genomes described here appears to be similar to that of other poxviral genomes studied (Niles et al., 1986; Moss, 1990; Goebel et al., 1990). Specifically, the two leporipox viruses described here resemble vaccinia in that they lack substantial intergenic regions. In this respect fowlpox virus appears to be an exception in that intergenic regions up to 900 bp have been found in this portion of the genome. Fowlpox virus appears further to differ from SFV, MV and vaccinia in that or-f’s with homology to the latter three virus’ orf’s in this region are separated by orf s with no clear homology to SFV, MV or VV orf s. Open reading frames are often arranged head-to-head in the same or different reading frames, or head-to-tail, read in opposite directions. MV and SFV orfs are also similar, if not identical, in size and composition to their homologs in vaccinia. Comparing SFV and MV in this portion of their genomes, we find a surprising degree of conservation of structure, both at the levels of DNA and orf sequences.
I- SFV OrI 0.4 ends .-ISFV ori D-S ends ACGCGCACAATGTATCAGTGAcATTGAGGAGCAOeAG~GGA~~AA~ TCAAATTTTDTGTTTCTA~CGTCCATAOTRGACTCOTCT~~~~~~~ 9000 hCSGGGACAATGTATCAOTG~~=*~~~~~~~~~~~~~~~~~~~~~~~~~ ?ChAI1TTTTOTGT?TC’**~~~:CCATAG*ADACTCGTCTGGAfAACOA^UT 71*0 1.. MVOrf 09 LwdS -I MV orf c-e ends GCSTACGTCGGTTOTATAOAACAACGGATRTRACOORTAA~~ :^‘;TACGTCCIGTTG~AI‘~~,,~~~~~~~~~~~~*~~~*~~~~~~~~~~~~~~
ATACATACAOTCOCTRTARCGAGGSCAIATRACTCARCA AThCAPAChCTC~CTA*A*CGAGZGCA7r\TAACr’hhCR~:~~~‘17
GTACAUGTG~~‘CTG*C‘ATARCRCCCTTCCGCGTTAGAAR~~~*~~~~~~~~ fGTTATACTCCTT+ACrSG~*~~:~~~~~~~~~~~~~TTA7TGTAAA?GT26 ‘:l‘h(‘A’T’T’;TrT1TCh?~,,~~~,:~~(‘T1“‘Tr,TCTTRCiAhRIl”SA,~;C’?TITft mTTRTACTCCTTTAl’r\G~~~-,.,-,,~,.Ti‘7r’Ti”mnAAi7A7~~~*~,~‘~:T-~, Fig.
4 1:t1. 7380 /, /* ‘?lSO
3 (ccmtinued).
This is in part reflected by the fact that most SFV orf’s do not differ by more than 10% in amino acid sequence from their homologs in MV. One of the pair is completely identical while others differ only in a few amino acids. Similarly large portions of the genomes from MV and SFV are identical in this area. This results in orf’s D-9 (SFV) and C-9 (MVI being identical. These orf’s encode a protein that, in vaccinia. is a virion membrane protein (Niles and Seto, 1988; Lee-Chen and Niles, 1988ab). By extension, its analogs in SFV and MV are probably membrane proteins as well. Perhaps its identity in these two viruses reflects a host range function or cellular tropism that both viruses share. The differences between these two orf’s and their vaccinia homofog could reflect narrower ranges of infectivity found in the two leporipoxviruses. Perhaps by chance, this protein also bears substantial homology to carbonic anhydrases from many species (Piatigorski and W&tow, 19891. Despite the impressive resembiance of MV and SFV orf’s, this area is not particularly conserved between these leporipox viruses on the one hand and vaccinia virus on the other: comparison shows that of the 12 orf’s in this area, this is one of the least conserved between the leporipox viruses and vaccinia. This orf is lacking in the corresponding region of fowlpox virus fTartaglia et al., 1990).
Because of this long stretch of DNA identity there is considerable identity in orf’s D-8 (SFV) and C-8 (MV) and D-10 and C-10. The former pair are homologous to a small subunit of DNA-dependent RNA polymerase in vaccinia. The latter are homologous to a vaccinia protein of unknown function that may be a DNA binding protein (E. Niles, personal communication). As with orf’s D-9 and C-9 compared to vaccinia D-9, these two pairs of leporipoxvirus orf’s are not particularly close to their homologs in vaccinia. The near-identi~ of SFV and MV sequences for orts D-8 and C-8 is of considerable interest because this orf lies adjacent to the area of the genorne where transfer of MV DNA to SFV dramatically alters virulence of resulting recombinant viruses (Strayer et al., 1990; Heard et al., 1990). The EarnHI “c” fragment of MV is critical to MV’s ability to infect lymphocytes and inhibit their proiiferation. When this MV DNA is transfected into fibroblasts infected with SFV, recombinant viruses result that are both capable of replicating lymphocytes and inducing the syndrome of disseminated tumors and severe immunologic suppression characteristic of MV (Strayer et al., 1988, 1990). To understand how this occurs, one must know the structure of the SFV DNA whose functions are superseded by those of the MV DNA, and which functions are at the same time inadequate to permit virus replication in lymphocytes and
128
13.00”
Fig. 2 (continued).
129 TABLE 1 Characteristics
of the 12 open reading frames contained in SFV’s BanrHI “D” fragment
The direction of the arrow indicates the direction of transcription. Homologies with MV, fowlpox virus (FPV) and vaccinia (WI orfs are indicated. The sizes of homologous orfs from these other viruses in kDa are indicated in brackets for comparison. Nomenclature for orfs is taken from Niles et al. (19861 and Tartaglia et al. (1990). Functional information available is summarized by Lee-Chen and Niles (1988a.b). * These orf’s in SFV and MV are homologous to vaccinia orf’s at variable distances from the NH? termini of the VV orfs. Therefore, since these are the lateralmost orfs of the SFV and MV resthction fragments. the orf’s as observed in SFV and MV probably do not include the full transcripts. s This orf is called D-72 in GenBank but ORFg in Niles et al. (1986). n Carbonic anhydrases homologous to D-9, in decreasing order of homology: rabbit CA II, human CA II, rabbit CA 1, mouse CA 1, chicken CA 11, mouse CA II, horse CA III, human CA III. Sheep CA II. bovine CA II, sheep, CA VI. rhesus CA I, horse CA I, human CA I. ORF number
Location
Length
Size
Homologies
D-l
536
--t 2396
620
72,135
D-2
?3Gl
+- 2790
143
16,554
D-3
2797 -+3510
241
27,904
D-4
3519
-+4173
218
25,540
D-5
4209 + 6567
786
90,791
D-6
5499
c- 5931
144
15,988
D-7
6566
-) 8471
635
73,205
D-8
8504
+8963
163
18,391
D-9
8967 -9822
285
32.346
D-10
9833
--t 10,487
218
25,217
D-11
10.486 -t 11,266
260
30,425
D-12
11,266 + 13,021
58.5
67,403
G-l (MVl[17.5]; D-l (VV1[96.7]* Subunit, mRNA capping enzyme C-2 (MV) [ 16.71;D-2 (VVI [16.9] Virion protein C-3 tMV)[27,8]; D-3 (VV, 128.01 Essential protein, role not known C-4 (MV) [25.3]; D-4 (VV) [X5.0] Function unknown C-5 tMVl[90.5]; D-5 (WI 190.31 FPD-5 (FPV) [92.7] DNA binding protein, unknown function C-6 (MV) [ 15.91;ORFgs (VV1[8.6] Function unknown C-7 (MV) 173.21;D-6 (VV) [68.4] FPD-6 (FPV) [6X7] Early transcription factor subunit C-8 (MVI [18.4]; D-7 (VV) [17.9] FDP-7 (FPV) [18.0] Subunit, DNA-dependent RNA polymerase C-9 (MV) [32.3]; D-S WV) [35.4] Virion membrane protein, CA’ss C-10 (MVI ]25.2]: D-9 (VV1[25.0] Possible DNA binding protein C-11 (MV) [30.4]; D-10 (VVl[28.9] Function unknown C-14 (MV) I42.41;D-11 (VV) [68]* Nucleoside triphosphatase
suppression of lymphocyte function. This DNA must be compared to that of the corresponding MV DNA involved. Interestingly, it appears that the C-7 orf of MV, which corresponds to orf D-7 from SFV, is the principal orf involved in this transfer of virulence (Heard et al., 1990; Strayer et al., 1991; D. Strayer, unpublished observations). These two open reading frames are relatively well conserved between the two viruses and are both
MVISFVNVIFPV
orf
identity
D8
09
80
% Identity 60 with SFV oti 40
&I % idenlfly (MV)zO
n
% identity (VV)
El
.% tdentity (FPV) 0 Dl
D2
D3
04
05
D6
D7
DlO
011
D12
SFV ORF
Fig. 3. Similarities between SFV, MV, vaccinia and fowlpox virus orf’a. The degree of identity between orf’s in MV, VV and FPV and their corresponding SFV BumHI “D” fragment orf’s is shown. It is displayed as the percentage identity for the amino acids in each poxvirus orf, compared to the corresponding amino acids of the homologous SFV orf. The name of the SFV orf is indicated on the bottom, and the percentage identity of MV orfs is shown in solid bars, of VV in diagonally lined bars and of FPV in shaded bars. Some orfs in SFV. MV and VV have no apparent homologs in FPV: some FPV orf’s lack homologs in the other viruses. In these circumstances, no FPV bar appears for the comparison with the SFV orf in question. ORF’s D-9 (SFV) and C-9 (MV) are identical.
similar to D-7 orf of vaccinia (Broyles et al., 1990). Vaccinia D-7 encodes a subunit of an early transcription factor. The apparent intimate involvement of a transcription factor in determination of at least some facets of virus virulence suggests that control of transcript levels, rather than the transcripts themselves, may be the key into the interactions of these two viruses with their host target cells (Strayer, 1992). These studies grew from an observation that considerable differences in virus virulence could reflect the activity of a small number of key genes. Transfer of a small portion of MV’s BarnHI “C” fragment of SFV alters the biological activity of resultant recombinant viruses dramatically. We report here the structural basis for this phenomenon. in the comparati~c sequences of the portions of the two viruses involved. Future work will hopefully allow us to localize specifically the amino acid difference(s) that determine this biological behavior and to understand the relationship between virulence and the structures that regulate it.
Acknowledgements The authors gratefully acknowledge personal communications about this work with Drs. Steven Broyles, Mark Bulier, Ed Niles and Chris Upton, The technical assistance of Kathleen O’Connor was helpful in this work. This work was supported by NIH Grant ~A44800.
131
References Ahlstrom. C.G. and Andrews. C.H. (1938) Fibroma virus infection in starved rabbits. J. Pathol. 47. 65-X6. Allison, A.C. (1966) Immune responses to Shope fibroma virus in adult and newborn rabbits. J. Natl. Cancer Inst. 36, 869-876. Allison, A.C. and Friedman, R.M. (1966) Effects of immunosuppressant on Shope rabbit fibroma. J. Natl. Cancer Inst. 36. 859-868. Bergman. S.. Jonsson. N. and Ahlstrom. C.G. (1962) Infectious fibroma in prednisolone-treated rabbits. Acta Pathol. Microbial. Stand. 55, 39-48. Broyles, S.S.. Li, J. and Fesler, B.S. (1990) Vaccinia virus gene encoding a component of the viral early transcription factor. J. Viral. 64: 1523-1529. Cabirac. G.F.. Strayer, D.S., Sell. S. and Leibowitz. J.L. (1985) Characterization, molecular cloning. and physical mapping of the Shope fibroma virus genome. Virology 143, 663-670. Clemmensen. J. (1939) The influence of Roentgen radiation on immunity to Shope fibroma virus. Am. J. Cancer 36, 378-385. Duran-Reynolds, F. (1940) Production of degenerative inflammatory or neoplastic effects in the newborn rabbit by the Shope fibroma virus. Yale J. Biol. Med. 13. 99-110. Duran-Reynolds, F. (1946) Immunological factors that influence the neoplastic effects of the rabbit fibroma virus. Cancer Res. 5, 25-39. Goebel. S.J., Johnson, G.P.. Perkus, M.E.. Davis, S.W., Winslow, J.P. and Paoletti. E. (1990) The complete DNA sequence of vaccinia virus. Virology 179, 247-266. Heard. H.K.. O’Connor. K. and Strayer, D. (1990) Molecular analysis of immunosuppression induced by virus replication in lymphocytes. J. Immunol., 144: 3992-3999. Hurst. E.W. (1964) The effect of cortisone and of h-mercaptopurine on the Shope fibroma. J. Pathol. Bacterial., 87. 29-37. Lee-Chen, G.-J. and Niles. E.J. (1988a) Transcription and translation mapping of the 13 genes in the vaccinia virus Hind111 D fragment. Virology 163. 52-63. Lee-Chen, G.-J. and Niles, E.G. (1988b) Map positions of the 5’ ends of eight mRNAs synthesized from the late genes in the vaccinia virus Hind111 D fragment. Virology 163, 80-92. Maniatis. T., Fritsch, E.F. and Sambrook. J. (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory. Cold Spring Harbor, N.Y. Moss. B. (1990) Poxviridae and their replication. In B.N. Fields, D.N. Knipe et al.. (Eds.), Virology. pp. 2079-2111. Raven Press, New York. Niles, E.G.. Condit, R.C., Caro. P., Davidson, K., Matusick, L. and Seto, J. (1986) Nucleotide sequence and genetic map of the 16.kb vaccinia virus Hind111 D fragment. Virology 153, 96-l 12. Niles, E.G. and Seto. J. (1988) Vaccinia virus gene D8 encodes a virion transmembrane protein. Virology 62, 3772-3778. Pearson, W.R. and Lipman, D.J. (1988) Improved tools for biological sequence comparison. Proc. Natl. Acad. Sci. USA 85. 2444-2448. Piatigorsky, J. and Wistow. G.J. (1989) Enzyme/crystallins: gene sharing as an evolutionary strategy. Cell 57. 197-199. Sell, S. and Scott, C.B. (1981) An immunohistologic study of Shope fibroma virus in rabbits: tumor rejection by cellular reaction in adult and progressive systemic reticuloendothehal endothelial infection in neonates. J. Natl. Cancer Inst., 66, 363-373. Shope, R.E. (1932a) A transmissible tumor-like condition of rabbits. J. Exp. Med. 56, 793-802. Shope. R.E. (1932b) A filterable virus causing a tumor-like condition in rabbits and its relationship to virus myxomatosum. J. Exp. Med., 56, 803-822. Smith, J.W.. Tevethia. S.S.. Levy, B.M. and Rawls, W.E. (1973) Comparative studies of host responses to Shope fibroma virus in adult and newborn rabbits. J. Natl. Cancer Inst. 50. 1529-1439. Smith, M.H.D. (1952) the BerryDedrick transformation of fibroma into myxoma with rabbit. Ann. N.Y. Acad Sci. 54, 1141-1152. Strayer, D.S., Skaletsky. E. and Sell, S. (1984) Strain differences in Shope fibroma virus: an immunopathologic study. Am. J. Pathol. 116, 342-358.
132 Strayer. D.S., Skaletsky, E., Cabirac. Cr., Sharp, P.A.. Corbeil, L.B., Sell. S. and Leihowitz, J.L. (1983a) Malignant rabbit fibroma virus causes secondary immunosuppression in rabbits. J. Immunol. 130. 399-404. Strayer, D.S.. Skaletsky. E., Leibowitz, J.L. and Dombrowski. J. (1987) Growth of malignant rabbit fibroma virus in lymphoid cells. Virology 158. 147-157. Strayer, D.S.. Mulloy, J. and Leibowitt. J.I.. (1988) Lymphocyte-virus interactions: identification of a restriction fragment permitting virus replication in lymphocytes. J. Immunol. 141. 2060-2067. Strayer, D.S.. Laybourn, K.A. and Heard, H.K. (1990) Determinants of the ability of malignant fibroma virus to induce tumor dissemination in vivo. Microb. Pathogen. 9. 173-189. Strayer, D.S., Jerng, H.H. and O’Connor. K. (1991) Sequence and analysis of a portion of the genomes of Shope fibroma virus and malignant rabbit fibroma virus that is important for viral replication in lymphocytes. Virology 185. 585-595. Strayer, D.S. (19921 Determinants of virus-related suppression of immune responses as observed during infection with an oncogenic poxvirus. Prog. Med. Virol., 39. X8-255. Tartaglia, J., Winslow, J., Goebel. S.. Johnson. G.P.. Taylor, J. and Paoletti, E. (1990) Nucleotide sequence analysis of a 10.5 kb Hind111 fragment of fowlpox virus: relatedness to the central portion of the vaccinia virus NindIII D region. J. gen. Viral. 71, 1517-1524. Upton, C.. Macen. J.L.. Maranchuk, R.A., DeLange, A.M. and McFadden. C. (1988) Tumorigenic poxviruses: fine analysis of the recombination junctions in malignant rabbit fibroma virus, a recombinant between Shope fibroma virus and myxoma virus. Virology_l66, 229-239. Upton, C., Stuart, D. and McFadden, G. ( 1991) Identification and DNA sequence of the large subunit of the capping enzyme from Shope fibroma virus. Virology 183, 773-777. Yuill, T.M. and Hanson, R.P. (19641 Infection of suckling cottontail rabbits with Shope’s fibroma virus. Proc. Sot. Exp. Biol. Med. 117, 376-380. Woodroofe, G.M. and Fenner. F. (1965) Viruses of the myoma-fibroma subgroup of poxviruses. I. Plaque production in cultured cells. plaque-reduction tests and cross-protection tests in rabbits. Aust. J. Exp. Biol. Med. Sci. 43. 123-142.
117
VIRUS 00814
Sequence and analysis of the BamHI “D” fragment of Shope fibroma virus: comparison with similar regions of related poxviruses David S. Strayer and Henry H. Jerng Department of Padlology and Cell Biology, Thomas Jefferson Utkersity Medical College, Philadelphia, PA, USA (Received 2 April 1992; revision received 26 May 1992: accepted 28 May 1992)
Summary Differences observed in the virulence of two related Ieporipoxviruses are closely tied to a particular region of their genomes. For the virulent poxvirus of this pair, malignant rabbit fibroma virus (MV), this region is the BanaH “C” fragment. which is 10.7 kb. For the avirulent poxvirus, Shope fibroma virus, SFV, this region is the corresponding BarnHI “D” fragment, which is 13.1 kb. As part of our attempt to understand the virulence of these two viruses, we sequenced these two DNA fragments. The sequence for the BarnHI “C” fragment of MV is reported elsewhere (Strayer et al., 1991). We report here the sequence for SFV’s BanzHI “D” fragment and resultant open reading frames, and compare both DNA and open reading frame structures to those of MV and other known poxviruses. The BamHI “D” fragment of SFV contains 12 open reading frames of 100 amino acids or more, arranged similarly to orf’s in MV and vaccinia. Striking similarities between SFV and MV are seen in certain parts of this restriction fragment, including substantial stretches of DNA in which the two viruses are identical. Clear homologies exist between these leporipox virus genomes and those of other related poxviruses. To understand the pathogenesis of virus infection, one must appreciate the structure of those viral genes that play important roles in infection. Malignant rabbit fibroma virus; Shope fibroma virus; Poxvirus; DNA sequence
Correspondence fcx D.S. Strayer, Deparrm~nt of Pathology and Cell Biology, Thomas Jefferson University Medical College, 1020 Locust Street, Philadelphia. PA 19107, USA. Fax: (13(215)-953-2218. Abbreviations: bp, base pairs; FPV, fowlpox virus; MV, malignant rabbit fibroma virus; MYX, rabbit myxoma virus; orf, open reading frame: SFV, Shope fibroma virus: W, vaccinia virus.
Introduction
Shope fibroma virus (SFV) was first isolated and identified in 1932 in wild rabbits in New Jersey (Shape, 1932a). At the time it was appreciated that this poxvirus was responsible for small, benign tumors that waxed and waned in rabbits over the course of several weeks. The relationship between SFV and its more virulent relative, rabbit myxoma virus (MYX), was established: prior infection with SFV protected rabbits against MYX infection (Shape, 1932b: Woodroofe and Fenner, 1965). Shortly thereafter a series of reports suggested that SFV behaved differently in very young or immunologically suppressed recipients (Ahlstrom and Andrews, 1938; Clemmensen, 1939; Duran-Reynolds, 1940; Duran-Reynolds, 1946; Bergmann et al.. 1962; Yuill and Hanson, 1964; Hurst, 1964; Allison and Friedman. 1966; Allison, 1966; Smith et al., 1973: Sell and Scott, 1981). In such animals, SFV infection caused disseminated tumors and often death. These data imply that SFV’s avirulence in normal adults is mostly a function of host immune responsiveness. Thus, differences in virulence between MYX and- SFV could be more apparent than real, SFV always producing disseminated lethal tumors in the “correct” host. We recentIy re-examined these findings. Using clonally purified stocks of SFV, we found that almost all strains of SFV caused only localized disease in steroidtreated or newborn rabbits. Only Boerlage strain SFV produced disseminated infection with necrosis in lymphoid organs in immunosuppressed and newborn rabbits. This strain, however, did not cause disseminated tumors and was not lethal (Strayer et al., 1984). Analysis of the several isolates of SFV showed that Boerlage strain SFV was genetically closer to MYX than were the other SFV strains (G. Cabirac et al., unpublished). Growing MYX and SFV together in vitro resulted in recombinant viruses of intermediate virulence (Smith, 1952). These observations indicate that SFV and MYX are sufficiently genetically related to produce recombinant viruses by co-culturing and underscore the role of viral gene structure in the course of these infections. We have been examining the virulence of malignant rabbit fibroma virus (MV). MV is a naturally occurring recombinant between SFV and MYX (Strayer et al., 19X3& the point of recombination occurred at the periphery of the virus genome. MV’s central sequences are from MYX while its terminal inverted repeat (TIR) region and a short span of unique sequences come from SFV (Upton et al., 1988). Clinically. MV is virulent, inducing disseminated tumors, severe gram-negative infection and death within 12 days. An important part of MV’s clinical behavior involves its ability to replicate in lymphocytes and suppress immune function (Strayer et al., 1987). SFV does neither of these. When MV’s 10.7 kb BamHI "c" fragment (hereafter, “C” fragment) is transferred to SFV, recombinant viruses so generated replicate in lymphocytes, suppress immune function and recapitulate most of MV’s aggressive phenotype
119
(Strayer et al., 1988, 1990). This fragment is located in the middle of the unique sequences of MV. Progressively smaller subfragments of the “C” fragment may transfer these characteristics. Two subfragments that overlap for about 700 bp are capable of transferring much of MV’s virulence to SFV: a 3.6 kb NdeI fragment and a 1.9 kb HirzcII fragment (Heard et al., 1990). The sequence of this portion of MV’s genome has been reported (Strayer et al., 1991). For Shope fibroma virus, the restriction fragment corresponding to MV’s “C” fragment is the 13.1 BamHI “D” fragment (hereafter “D” fragment). This fragment extends beyond both ends of MV’s “C” fragment. Both the C and D fragments are also homologous to the Hind111 “D” fragment of vaccinia virus (WR strain, Niles et al., 1986) and to a 10.5 kb Hind111 fragment of fowlpox virus (Tartaglia et al., 1990). Because of the importance of this region of the viral genome in the determination of virulence, we report here the sequence of SFV’s BamHI “D” fragment in its entirety and compare it to sequences of related poxviruses in this area. Small portions of this sequence have been reported elsewhere by others (Upton et al., 1991) and by ourselves (Strayer et al., 1991).
Materials
and Methods
SFVh BamHI “D” fragmerlt
The BamHI “D” fragment of SFV was obtained from cloned material generated by G. Cabirac (Cabirac, et al., 1987). A restriction map was prepared using HincII, NdeI, EcoRI, Xl101 and Hind111 (Fig. 1).
2
0
4
6
I
I
S
I
'---)z~iT~
I
7+3
4-6
9-b
10
12
I
I
I‘lkh
10 +11 + 9e
-
12
Hire
II
Em
RI
Nde
I
Hind Ill Xho I
Fig. 1. Below: restriction map of SFV BumHI “D” fragment from which sequencing was done. In this illustration. the enzymes HincII, NdeI, EcoRI. 27~01 and Hind111 are used to subdivide the 13.1 kb BarnHI fragment. Approximate sizes are shown in kb within the respective fragment boxes. Above: open reading frame organization for SFV’s BamHI “D” fragment. orf’s read from left to right are D-l. -3. -4. -5. -7, -8. -10, and -11. The others are read from right to left.
The “0” fragment was digested with the above enzymes and cloned either as singly cut or doubly cut restriction subfragments into sequencing vectors of the pGEM series (Promega). Sequencing was done by the dideoxy method (Maniatis, 1982) using first the SP6 and T7 promoters of the pGEM plasmids, plus Sequenase (USBC). (Only Sequenase was used successfully in these sequencing reactions. We tried Sequenase 11 (USBC) but were unable to obtain sequence data with it.1 Sequences further into the restriction fragments than could be obtained using the plasmid primers were sequenced using oligonucleotide primers synthesized to complement viral sequences about 50 bp upstream of the end of the previous sequence. Sequences at junctions of restriction fragments were obtained by sequencing overlapping restriction fragments. The SFV DNA sequence for this fragment has been reported to GenBank and carries accession number M74532. The sequence of MV’s BanrHI “C” fragment has accession number M32743. Sequence analysis
Sequences were assembled and analyzed using a Macintosh II computer (Apple Corp.1 with DNA Inspector IIe software (Textco, Inc.). Open reading frame (orfl recognition was performed using DNA Inspector Be. Databank
Protein and DNA sequences were compared to sequences available in GenBank via modem, using homology algorithms developed and described elsewhere (Pearson and Lipman, 1988).
Results Sequence of SFV BamHI “D” fragment
SFV’s BavrzHI “D” fragment is a 13.088 bp restriction fragment. Its sequence is shown in Fig. 2. Locations and directions of transcription for open reading frames are indicated. The organization of orf’s in SFV is very simiIar to that of other pox viruses studied, including vaccinia virus (Niles et al., 19861, malignant rabbit fibroma virus (Strayer et al, 1991) and fowlpox virus (Tartaglia et al., 19901. There are no long spaces between adjacent orfs. Often orf’s read in the same direction in different frames, or in different directions overlap each other by several bases. Open rending ~framcs
Open reading frames located in the fragment and their similarity to orf’s in MV. VV and FPV are shown in Table 1 and illustrated in Fig. 3. There are 12 orf’s in the SFV “D’” fragment. orfs D-l, -3, -4, -5, -7, -8, -10 and -11 are read left to right. The others (D-2, -6, -9 and -121 are read right to left. Except for D-12, these orf’s correspond to like-numbered orf’s in MV’s Barn “C” fragment.
Fig. 7. (Continued on pp. 122-128.) Sequence of SFV’s BumHI “D” fragment and comparison with MV’s BumHI “C” fragment. The nucleotide sequences of these two restriction fragments are shown and aligned. Positions that had to be skipped to produce alignment are shown as “.“. The larger SFV fragment is shown above the smaller MV fragment in those areas that correspond, and the numbering in plain type is indicated for the SFV sequence. Numbering for the MV sequence is shown in bold faced type. Individual lines are arranged to provide 100 SFV bases/line. Beginning and ending points for the SFV and MV open reading frames are indicated, with initial ATG’s underlined and arrows indicating the direction of transcription. The notations for SFV orf’s are above the corresponding sequences: those for MV in italics below the sequences.
Open reading frames in SFV and MV are extremely similar in size and amino acid composition. Both sets of orf’s resemble those of vaccinia virus as well as some of the fowlpox orfs that localize to the corresponding portion of fowlpox virus. The resemblance to vaccinia and fowlpox viruses is much less than that to MV. Where there is a substantial difference in orf length among the several viruses, SFV and MV both differ from vaccinia and fowlpox and resemble each other. Based on functional analyses performed for some of the open reading frames in vaccinia’s Hind111 “D” fragment, it is possible to infer the functions of SFV’s orf’s. These homologies and corresponding inferences are listed in Table 1.
As indicated above, SFV is genetically closely related to MV although their biological behavior is greatly different. As MV’s BarnHI “C” fragment is a major factor in the differential virulence of MV and SFV, we compared DNA sequences for SFV and MV in this region, covering SFV’s 13.1 kb Banz “D” fragment and MV’s 10.7 kb i?anr “C” fragment. The two sequences are very similar (Fig. 2). Locations of orf’s are identical for both viruses, save for two small orf’s present in MV but not SFV. and located near the right side of the corresponding fragment. We examined the differences between SFV and MV in this portion of the genome relative to position in the “C” fragment. At no point do the two viruses differ in more than 159%of their bases. However, substantial stretches of DNA of these two viruses are identical. This distribution is non-random (P +z 0.01). There are no base differences between SFV and MV for a stretch of 326 bp between 5301 and 5626 (SFV). This area is entirely within orf’s D-5 and C-5 for SFV and MV respectively, and also includes the CO, terminus of orfs D-6 and C-6. These latter orf’s are entirely contained within D-5 and C-5, and are transcribed in the opposite direction.
In addition, there are no differences between SFV and MV for 209 bases (8504-8712 for SFV) at the very beginning of orfs D-S and C-8. This is followed by a short stretch of 55 bp in which there are 6 differences between the two viruses resulting in two amino acid changes (I -+ V and T + I), and in turn by 1481 bp (886%10,348) without a single divergent base between the two viruses. This portion of the genome includes the carboxyl half of orf’s D-8 and C-8 (probably a subunit of DNA-dependent RNA polymerase), the entire orf’s D-9 and C-9 (probably a virion membrane protein) and almost all of orfs D-10 and C-10 (function unknown). Interestingly, none of these areas of identity between SFV and MV correspond to an area of strong homology between either of these two viruses and vaccinia. Open reading frames in this area are strongly divergent between SFV and MV on the one hand and vaccinia on the other. In those instances where orf’s for fowlpox virus have been determined and correspond to SFV, MV or VV, FPV orf’s differ from the two Ieporipox viruses more than do those of VV (Tartaglia et al., 1990).
124
Discussion
The sequence of the SFV BanrH1 “D” fragment closely resembles that of the MV BarnHI “C” fragment, is similar to vaccinia’s Hind111 “D” fragment (Niles et al., 1986) and relates somewhat more distantly to the genome of fowlpox virus as reported by Tartaglia et al. (1990). Upton and co-workers (19911, studying SFV’s capping enzyme, have sequenced the left-most portion of Kasza strain SFV. Their sequence corresponds to the first 3.5 kb of the Hi~rdIII “D” fragment, and is substantiaIly in agreement with that reported here. As shown above, the area of homology to the entire MV “c” fragment of MV is contained within SFV’s BamHI “D” fragment. The 13,088 bp of SFV’s “D” fragment is homologous to a portion of vaccinia’s 16.1 kb f?indIII “D” fragment that is completely contained within the latter restriction fragment. Open reading frame D-l from SFV corresponds to orf D-l from vaccinia with the exception that the latter orf contains 210 amino acids at the NH,-terminus not present in D-l from SFV. This difference undoubtedly represents the fact that the former orf begins near the 5’ end of the Hind111 “D” fragment of VV and that the latter orf begins near the 5’ end of SFV’s BumHI ‘“D” fragment and does not encompass the entire 5’ end of the orf.
An analogous situation obtains with SFV orf D-12, compared to W orf D-11. Both of these orf’s are read right to left. The latter orf contains 46 amino acids at the NH,-terminus that are not in SFV’s D-12. Thus, the orf from SFV probably begins 5’ to the right side of the BamHI “D” fragment of SFV and does not reflect the true size of the transcript. The organization of leporipoxviral genomes described here appears to be similar to that of other poxviral genomes studied (Niles et al., 1986; Moss, 1990; Goebel et al., 1990). Specifically, the two leporipox viruses described here resemble vaccinia in that they lack substantial intergenic regions. In this respect fowlpox virus appears to be an exception in that intergenic regions up to 900 bp have been found in this portion of the genome. Fowlpox virus appears further to differ from SFV, MV and vaccinia in that or-f’s with homology to the latter three virus’ orf’s in this region are separated by orf s with no clear homology to SFV, MV or VV orf s. Open reading frames are often arranged head-to-head in the same or different reading frames, or head-to-tail, read in opposite directions. MV and SFV orfs are also similar, if not identical, in size and composition to their homologs in vaccinia. Comparing SFV and MV in this portion of their genomes, we find a surprising degree of conservation of structure, both at the levels of DNA and orf sequences.
I- SFV OrI 0.4 ends .-ISFV ori D-S ends ACGCGCACAATGTATCAGTGAcATTGAGGAGCAOeAG~GGA~~AA~ TCAAATTTTDTGTTTCTA~CGTCCATAOTRGACTCOTCT~~~~~~~ 9000 hCSGGGACAATGTATCAOTG~~=*~~~~~~~~~~~~~~~~~~~~~~~~~ ?ChAI1TTTTOTGT?TC’**~~~:CCATAG*ADACTCGTCTGGAfAACOA^UT 71*0 1.. MVOrf 09 LwdS -I MV orf c-e ends GCSTACGTCGGTTOTATAOAACAACGGATRTRACOORTAA~~ :^‘;TACGTCCIGTTG~AI‘~~,,~~~~~~~~~~~~*~~~*~~~~~~~~~~~~~~
ATACATACAOTCOCTRTARCGAGGSCAIATRACTCARCA AThCAPAChCTC~CTA*A*CGAGZGCA7r\TAACr’hhCR~:~~~‘17
GTACAUGTG~~‘CTG*C‘ATARCRCCCTTCCGCGTTAGAAR~~~*~~~~~~~~ fGTTATACTCCTT+ACrSG~*~~:~~~~~~~~~~~~~TTA7TGTAAA?GT26 ‘:l‘h(‘A’T’T’;TrT1TCh?~,,~~~,:~~(‘T1“‘Tr,TCTTRCiAhRIl”SA,~;C’?TITft mTTRTACTCCTTTAl’r\G~~~-,.,-,,~,.Ti‘7r’Ti”mnAAi7A7~~~*~,~‘~:T-~, Fig.
4 1:t1. 7380 /, /* ‘?lSO
3 (ccmtinued).
This is in part reflected by the fact that most SFV orf’s do not differ by more than 10% in amino acid sequence from their homologs in MV. One of the pair is completely identical while others differ only in a few amino acids. Similarly large portions of the genomes from MV and SFV are identical in this area. This results in orf’s D-9 (SFV) and C-9 (MVI being identical. These orf’s encode a protein that, in vaccinia. is a virion membrane protein (Niles and Seto, 1988; Lee-Chen and Niles, 1988ab). By extension, its analogs in SFV and MV are probably membrane proteins as well. Perhaps its identity in these two viruses reflects a host range function or cellular tropism that both viruses share. The differences between these two orf’s and their vaccinia homofog could reflect narrower ranges of infectivity found in the two leporipoxviruses. Perhaps by chance, this protein also bears substantial homology to carbonic anhydrases from many species (Piatigorski and W&tow, 19891. Despite the impressive resembiance of MV and SFV orf’s, this area is not particularly conserved between these leporipox viruses on the one hand and vaccinia virus on the other: comparison shows that of the 12 orf’s in this area, this is one of the least conserved between the leporipox viruses and vaccinia. This orf is lacking in the corresponding region of fowlpox virus fTartaglia et al., 1990).
Because of this long stretch of DNA identity there is considerable identity in orf’s D-8 (SFV) and C-8 (MV) and D-10 and C-10. The former pair are homologous to a small subunit of DNA-dependent RNA polymerase in vaccinia. The latter are homologous to a vaccinia protein of unknown function that may be a DNA binding protein (E. Niles, personal communication). As with orf’s D-9 and C-9 compared to vaccinia D-9, these two pairs of leporipoxvirus orf’s are not particularly close to their homologs in vaccinia. The near-identi~ of SFV and MV sequences for orts D-8 and C-8 is of considerable interest because this orf lies adjacent to the area of the genorne where transfer of MV DNA to SFV dramatically alters virulence of resulting recombinant viruses (Strayer et al., 1990; Heard et al., 1990). The EarnHI “c” fragment of MV is critical to MV’s ability to infect lymphocytes and inhibit their proiiferation. When this MV DNA is transfected into fibroblasts infected with SFV, recombinant viruses result that are both capable of replicating lymphocytes and inducing the syndrome of disseminated tumors and severe immunologic suppression characteristic of MV (Strayer et al., 1988, 1990). To understand how this occurs, one must know the structure of the SFV DNA whose functions are superseded by those of the MV DNA, and which functions are at the same time inadequate to permit virus replication in lymphocytes and
128
13.00”
Fig. 2 (continued).
129 TABLE 1 Characteristics
of the 12 open reading frames contained in SFV’s BanrHI “D” fragment
The direction of the arrow indicates the direction of transcription. Homologies with MV, fowlpox virus (FPV) and vaccinia (WI orfs are indicated. The sizes of homologous orfs from these other viruses in kDa are indicated in brackets for comparison. Nomenclature for orfs is taken from Niles et al. (19861 and Tartaglia et al. (1990). Functional information available is summarized by Lee-Chen and Niles (1988a.b). * These orf’s in SFV and MV are homologous to vaccinia orf’s at variable distances from the NH? termini of the VV orfs. Therefore, since these are the lateralmost orfs of the SFV and MV resthction fragments. the orf’s as observed in SFV and MV probably do not include the full transcripts. s This orf is called D-72 in GenBank but ORFg in Niles et al. (1986). n Carbonic anhydrases homologous to D-9, in decreasing order of homology: rabbit CA II, human CA II, rabbit CA 1, mouse CA 1, chicken CA 11, mouse CA II, horse CA III, human CA III. Sheep CA II. bovine CA II, sheep, CA VI. rhesus CA I, horse CA I, human CA I. ORF number
Location
Length
Size
Homologies
D-l
536
--t 2396
620
72,135
D-2
?3Gl
+- 2790
143
16,554
D-3
2797 -+3510
241
27,904
D-4
3519
-+4173
218
25,540
D-5
4209 + 6567
786
90,791
D-6
5499
c- 5931
144
15,988
D-7
6566
-) 8471
635
73,205
D-8
8504
+8963
163
18,391
D-9
8967 -9822
285
32.346
D-10
9833
--t 10,487
218
25,217
D-11
10.486 -t 11,266
260
30,425
D-12
11,266 + 13,021
58.5
67,403
G-l (MVl[17.5]; D-l (VV1[96.7]* Subunit, mRNA capping enzyme C-2 (MV) [ 16.71;D-2 (VVI [16.9] Virion protein C-3 tMV)[27,8]; D-3 (VV, 128.01 Essential protein, role not known C-4 (MV) [25.3]; D-4 (VV) [X5.0] Function unknown C-5 tMVl[90.5]; D-5 (WI 190.31 FPD-5 (FPV) [92.7] DNA binding protein, unknown function C-6 (MV) [ 15.91;ORFgs (VV1[8.6] Function unknown C-7 (MV) 173.21;D-6 (VV) [68.4] FPD-6 (FPV) [6X7] Early transcription factor subunit C-8 (MVI [18.4]; D-7 (VV) [17.9] FDP-7 (FPV) [18.0] Subunit, DNA-dependent RNA polymerase C-9 (MV) [32.3]; D-S WV) [35.4] Virion membrane protein, CA’ss C-10 (MVI ]25.2]: D-9 (VV1[25.0] Possible DNA binding protein C-11 (MV) [30.4]; D-10 (VVl[28.9] Function unknown C-14 (MV) I42.41;D-11 (VV) [68]* Nucleoside triphosphatase
suppression of lymphocyte function. This DNA must be compared to that of the corresponding MV DNA involved. Interestingly, it appears that the C-7 orf of MV, which corresponds to orf D-7 from SFV, is the principal orf involved in this transfer of virulence (Heard et al., 1990; Strayer et al., 1991; D. Strayer, unpublished observations). These two open reading frames are relatively well conserved between the two viruses and are both
MVISFVNVIFPV
orf
identity
D8
09
80
% Identity 60 with SFV oti 40
&I % idenlfly (MV)zO
n
% identity (VV)
El
.% tdentity (FPV) 0 Dl
D2
D3
04
05
D6
D7
DlO
011
D12
SFV ORF
Fig. 3. Similarities between SFV, MV, vaccinia and fowlpox virus orf’a. The degree of identity between orf’s in MV, VV and FPV and their corresponding SFV BumHI “D” fragment orf’s is shown. It is displayed as the percentage identity for the amino acids in each poxvirus orf, compared to the corresponding amino acids of the homologous SFV orf. The name of the SFV orf is indicated on the bottom, and the percentage identity of MV orfs is shown in solid bars, of VV in diagonally lined bars and of FPV in shaded bars. Some orfs in SFV. MV and VV have no apparent homologs in FPV: some FPV orf’s lack homologs in the other viruses. In these circumstances, no FPV bar appears for the comparison with the SFV orf in question. ORF’s D-9 (SFV) and C-9 (MV) are identical.
similar to D-7 orf of vaccinia (Broyles et al., 1990). Vaccinia D-7 encodes a subunit of an early transcription factor. The apparent intimate involvement of a transcription factor in determination of at least some facets of virus virulence suggests that control of transcript levels, rather than the transcripts themselves, may be the key into the interactions of these two viruses with their host target cells (Strayer, 1992). These studies grew from an observation that considerable differences in virus virulence could reflect the activity of a small number of key genes. Transfer of a small portion of MV’s BarnHI “C” fragment of SFV alters the biological activity of resultant recombinant viruses dramatically. We report here the structural basis for this phenomenon. in the comparati~c sequences of the portions of the two viruses involved. Future work will hopefully allow us to localize specifically the amino acid difference(s) that determine this biological behavior and to understand the relationship between virulence and the structures that regulate it.
Acknowledgements The authors gratefully acknowledge personal communications about this work with Drs. Steven Broyles, Mark Bulier, Ed Niles and Chris Upton, The technical assistance of Kathleen O’Connor was helpful in this work. This work was supported by NIH Grant ~A44800.
131
References Ahlstrom. C.G. and Andrews. C.H. (1938) Fibroma virus infection in starved rabbits. J. Pathol. 47. 65-X6. Allison, A.C. (1966) Immune responses to Shope fibroma virus in adult and newborn rabbits. J. Natl. Cancer Inst. 36, 869-876. Allison, A.C. and Friedman, R.M. (1966) Effects of immunosuppressant on Shope rabbit fibroma. J. Natl. Cancer Inst. 36. 859-868. Bergman. S.. Jonsson. N. and Ahlstrom. C.G. (1962) Infectious fibroma in prednisolone-treated rabbits. Acta Pathol. Microbial. Stand. 55, 39-48. Broyles, S.S.. Li, J. and Fesler, B.S. (1990) Vaccinia virus gene encoding a component of the viral early transcription factor. J. Viral. 64: 1523-1529. Cabirac. G.F.. Strayer, D.S., Sell. S. and Leibowitz. J.L. (1985) Characterization, molecular cloning. and physical mapping of the Shope fibroma virus genome. Virology 143, 663-670. Clemmensen. J. (1939) The influence of Roentgen radiation on immunity to Shope fibroma virus. Am. J. Cancer 36, 378-385. Duran-Reynolds, F. (1940) Production of degenerative inflammatory or neoplastic effects in the newborn rabbit by the Shope fibroma virus. Yale J. Biol. Med. 13. 99-110. Duran-Reynolds, F. (1946) Immunological factors that influence the neoplastic effects of the rabbit fibroma virus. Cancer Res. 5, 25-39. Goebel. S.J., Johnson, G.P.. Perkus, M.E.. Davis, S.W., Winslow, J.P. and Paoletti. E. (1990) The complete DNA sequence of vaccinia virus. Virology 179, 247-266. Heard. H.K.. O’Connor. K. and Strayer, D. (1990) Molecular analysis of immunosuppression induced by virus replication in lymphocytes. J. Immunol., 144: 3992-3999. Hurst. E.W. (1964) The effect of cortisone and of h-mercaptopurine on the Shope fibroma. J. Pathol. Bacterial., 87. 29-37. Lee-Chen, G.-J. and Niles. E.J. (1988a) Transcription and translation mapping of the 13 genes in the vaccinia virus Hind111 D fragment. Virology 163. 52-63. Lee-Chen, G.-J. and Niles, E.G. (1988b) Map positions of the 5’ ends of eight mRNAs synthesized from the late genes in the vaccinia virus Hind111 D fragment. Virology 163, 80-92. Maniatis. T., Fritsch, E.F. and Sambrook. J. (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory. Cold Spring Harbor, N.Y. Moss. B. (1990) Poxviridae and their replication. In B.N. Fields, D.N. Knipe et al.. (Eds.), Virology. pp. 2079-2111. Raven Press, New York. Niles, E.G.. Condit, R.C., Caro. P., Davidson, K., Matusick, L. and Seto, J. (1986) Nucleotide sequence and genetic map of the 16.kb vaccinia virus Hind111 D fragment. Virology 153, 96-l 12. Niles, E.G. and Seto. J. (1988) Vaccinia virus gene D8 encodes a virion transmembrane protein. Virology 62, 3772-3778. Pearson, W.R. and Lipman, D.J. (1988) Improved tools for biological sequence comparison. Proc. Natl. Acad. Sci. USA 85. 2444-2448. Piatigorsky, J. and Wistow. G.J. (1989) Enzyme/crystallins: gene sharing as an evolutionary strategy. Cell 57. 197-199. Sell, S. and Scott, C.B. (1981) An immunohistologic study of Shope fibroma virus in rabbits: tumor rejection by cellular reaction in adult and progressive systemic reticuloendothehal endothelial infection in neonates. J. Natl. Cancer Inst., 66, 363-373. Shope, R.E. (1932a) A transmissible tumor-like condition of rabbits. J. Exp. Med. 56, 793-802. Shope. R.E. (1932b) A filterable virus causing a tumor-like condition in rabbits and its relationship to virus myxomatosum. J. Exp. Med., 56, 803-822. Smith, J.W.. Tevethia. S.S.. Levy, B.M. and Rawls, W.E. (1973) Comparative studies of host responses to Shope fibroma virus in adult and newborn rabbits. J. Natl. Cancer Inst. 50. 1529-1439. Smith, M.H.D. (1952) the BerryDedrick transformation of fibroma into myxoma with rabbit. Ann. N.Y. Acad Sci. 54, 1141-1152. Strayer, D.S., Skaletsky. E. and Sell, S. (1984) Strain differences in Shope fibroma virus: an immunopathologic study. Am. J. Pathol. 116, 342-358.
132 Strayer. D.S., Skaletsky, E., Cabirac. Cr., Sharp, P.A.. Corbeil, L.B., Sell. S. and Leihowitz, J.L. (1983a) Malignant rabbit fibroma virus causes secondary immunosuppression in rabbits. J. Immunol. 130. 399-404. Strayer, D.S.. Skaletsky. E., Leibowitz, J.L. and Dombrowski. J. (1987) Growth of malignant rabbit fibroma virus in lymphoid cells. Virology 158. 147-157. Strayer, D.S.. Mulloy, J. and Leibowitt. J.I.. (1988) Lymphocyte-virus interactions: identification of a restriction fragment permitting virus replication in lymphocytes. J. Immunol. 141. 2060-2067. Strayer, D.S.. Laybourn, K.A. and Heard, H.K. (1990) Determinants of the ability of malignant fibroma virus to induce tumor dissemination in vivo. Microb. Pathogen. 9. 173-189. Strayer, D.S., Jerng, H.H. and O’Connor. K. (1991) Sequence and analysis of a portion of the genomes of Shope fibroma virus and malignant rabbit fibroma virus that is important for viral replication in lymphocytes. Virology 185. 585-595. Strayer, D.S. (19921 Determinants of virus-related suppression of immune responses as observed during infection with an oncogenic poxvirus. Prog. Med. Virol., 39. X8-255. Tartaglia, J., Winslow, J., Goebel. S.. Johnson. G.P.. Taylor, J. and Paoletti, E. (1990) Nucleotide sequence analysis of a 10.5 kb Hind111 fragment of fowlpox virus: relatedness to the central portion of the vaccinia virus NindIII D region. J. gen. Viral. 71, 1517-1524. Upton, C.. Macen. J.L.. Maranchuk, R.A., DeLange, A.M. and McFadden. C. (1988) Tumorigenic poxviruses: fine analysis of the recombination junctions in malignant rabbit fibroma virus, a recombinant between Shope fibroma virus and myxoma virus. Virology_l66, 229-239. Upton, C., Stuart, D. and McFadden, G. ( 1991) Identification and DNA sequence of the large subunit of the capping enzyme from Shope fibroma virus. Virology 183, 773-777. Yuill, T.M. and Hanson, R.P. (19641 Infection of suckling cottontail rabbits with Shope’s fibroma virus. Proc. Sot. Exp. Biol. Med. 117, 376-380. Woodroofe, G.M. and Fenner. F. (1965) Viruses of the myoma-fibroma subgroup of poxviruses. I. Plaque production in cultured cells. plaque-reduction tests and cross-protection tests in rabbits. Aust. J. Exp. Biol. Med. Sci. 43. 123-142.