VIROLOGY 178,28 1~ 284 (1990) SHORT COMMUNICATIONS Multiplication of Beet Necrotic Yellow Vein Virus RNA 3 Lacking a 3’ Poly(A) Tail Is Accompanied by Reappearance of the Poly(A) Tail and a Novel Short U-Rich Tract Preceding It I. JUPIN, S. BOUZOUBAA, K. RICHARDS,’ G. JONARD, AND H. GUILLEY institut de Biologic Mole’culaire des Plantes du CNRS, Universik Louis Pasteur, 12 rue du G&&al Zimmer. 67084 Strasbourg C6dex. France Received February 15, 1990; accepted May 8, 1990 Beet necrotic yellow vein virus RNAs 1 and 2 but not RNAs 3 and 4 are required forviral multiplication in Chenopodium quinoa leaves. Elimination of the 3’ poly(A) tail from RNA 3 transcripts markedly attenuated their ability to be amplified when co-inoculated with RNAs 1 and 2 to this host. Successful multiplication of the tailless RNA 3 was accompanied by the reappearance of new 3’ poly(A) tails on the progeny. The evidence suggests that the newly acquired poly(A) sequence results from the action of a poly(A) polymerase rather than recombination with the homologous 3’ terminal domains of RNAs 1 or 2. An unexpected feature of these progeny RNA 3 molecules was the presence of a novel short heterogenous U-rich tract separating the poly(A) tail from the 3’ end of the heteropolymeric RNA 3 sequence proper. 2) 1990 Academic Press. Inc. Beet necrotic yellow vein virus (BNYVV) is a soil- borne virus with a quadripartite, plus-strand RNA ge- nome (1, 2). In naturally infected sugar beet roots all four genome components are invariably present and apparently essential (3, 4) but when the virus is inocu- lated to leaves of Chenopodium quinoa neither of the two smallest RNA components, RNAs 3 and 4, are re- quired for virus multiplication (5-8). The confinement, in leaves, of replication functions to the two largest BNYVV RNAs, RNAs 1 and 2, evidently makes C. qui- noa a useful host for localizing sequences on the smaller RNAs important in cis for in viva multiplication. A biologically active transcript of RNA 3 has been ob- tained by bacteriophage T7 RNA polymerase-directed run-off transcription of cloned viral cDNA (9). Else- where, we have mapped c&essential domains within RNA 3 by inoculating C. quinoa leaves with wild-type RNAs 1 and 2 along with RNA 3 transcripts bearing in- ternal deletions introduced at the cDNA level (10). In this paper we have used a similar approach to investi- gate the influence of the 3’ poly(A) tail of RNA 3 on its amplification. All four natural BNYVV RNAs carry poly(A) tracts at their 3’ extremities (7 I). Such 3’ poly(A) tails, which are also characteristic of comoviruses, potyviruses, and potexviruses (12), could help to protect the heteropoly- merit portion of the viral RNA sequence from exonu- cleolytic attack. A role in the RNA replication process is likewise possible. At its 3’terminus, t35, the previously described wild-type infectious RNA 3 transcript (9), bears a poly(A) tract of about 110 nt followed by a 12 ’ To whom requests for reprints should be addressed. nt nonviral sequence derived from the vector. Recent observations have shown that the extra nonviral se- quence at t35’s 3’ terminus is not present on progeny RNA isolated from infected leaves (unpublished obser- vations). By manipulating the transcnptton vector pB35, which contains the full-length wild-type RNA 3 cDNA insert (9), we have produced modified RNA 3 run-off transcripts with 16 A residues or no A residues at the 3’ terminus. In order to do this, pB35 was linearized with HindIll (which cuts in the polylinker downstream of the insert poly(A) tail) and then digested under mild conditions with exonuclease Ill according to the suppli- er’s (Stratagene) instructions followed by treatment with mung bean nuclease. The DNAwas recircularized by ligation in the presence of excess Pstl-Bglll linker (!%pCTGCAGATCT) and introduced into Escherichia co/i JM 10 1. Clones were screened for the presence of the linker and shortened poly(A) tracts by restriction en- zyme digest/on and sequence analysis. A plasmid, pB35A,, , bearing a 16 residue poly(A) tract followed by the Pstl site, was selected. Single-stranded PB~~A,~ DNA was prepared by infecting JM 101 bacteria trans- formed with PB~~A,~ with the helper phage VCSM13 (9) and the poly(A) tract was eliminated completely by site-directed deletion mutagenesis of single-stranded DNA using a synthetic oligodeoxyribonucleotide com- plementary to the 3’terminal 14 residues of the hetero- polymeric portion of RNA 3 and the Pstl-Bglll linker. Mutated DNA was obtained using a commercial kit (Amersham) based on the phosphorothioate method of Taylor and Eckstein (13) and complying with the suppli- er’s instructions. Note that in the resulting non-poly(A)- tailed plasmid, pB35A,, the 3’terminal C residue of the