Tal­mu­dic Ques­tion #104

Oh yes, I can! One Deltapro­teobac­terium comes pretty close to the re­quired spec­i­fi­ca­tions: Bdellovib­rio bac­te­ri­ovorus: it at­taches to and in­vades Gram-neg­a­tive bac­te­ria (E. coli, Pseudomonas), re­sides in the periplasm of its host, sucks nu­tri­ents, grows & repli­cates and fi­nally ly­ses the host to set its prog­eny free. This is a virus-like lifestyle. How­ever, Bdello ap­par­ently doesn't en­gage any host fac­tors for ba­sic fea­tures as e.g. tran­scrip­tion, trans­la­tion and repli­ca­tion, as most viruses do. And: it does not en­ter the cy­to­plasm of its host. In the lab, you can get Bdello mu­tants that grow prey-in­de­pen­dently – even ax­eni­cally – thus they 'repli­cate as cells'.
P.S. I as­sume that My­coplasma lab­o­ra­to­rium does not count as an­swer to ques­tion #104 since it was cre­ated ar­ti­fi­cially in the Craig Ven­ter lab.

A tem­per­ate phage repli­cates like a cell in lysogeny and like a virus when in­duced to its lytic cy­cle,

Yes — it would have to be a kind of cross be­tween a gi­ant virus (Mimi- or Pan­dora- like) and a my­coplasma. In some con­di­tions, it could in­fect host cells as a virus, with com­plete dis­so­ci­a­tion of the in­fect­ing par­ti­cle at some point af­ter en­try, then DNA repli­ca­tion and for­ma­tion of new in­fec­tious par­ti­cles. In other con­di­tions, the in­fec­tious par­ti­cle it­self could be­come ca­pa­ble of growth and di­vi­sion. In this case, the in­fec­tious par­ti­cle would have to be some­thing like a bac­te­r­ial en­dospore — meta­bol­i­cally in­ert it­self, but ca­pa­ble of pro­duc­ing a veg­e­ta­tive cell un­der the right con­di­tions. Of course, you would need a large genome to code for these two dif­fer­ent life cy­cles, but maybe 2Mb would be enough.
While such a beast may be a the­o­ret­i­cal pos­si­bil­ity, I am not sure it could ex­ist in vivo. Vi­ral lifestyles se­lect for rapid repli­ca­tion, and loss of use­less genes, so a few cy­cles of virus-style repli­ca­tion might se­lect for trimmed down genomes that had lost es­sen­tial genes for the cel­lu­lar repli­ca­tion phase. Maybe the only way this could work in na­ture is if the two repli­ca­tion cy­cles had to al­ter­nate. But that would also re­quire a rather unique set of en­vi­ron­men­tal con­di­tions. (Host cells pe­ri­od­i­cally be­com­ing very rare, thus fa­vor­ing a cel­lu­lar lifestyle, then nu­tri­ents be­com­ing very sparse, thus fa­vor­ing vi­ral in­fec­tion of host cells). The host would have to be some kind of au­totroph, I guess, and once its pop­u­la­tion had been erad­i­cated by the vi­ral rec­pli­ca­tion cy­cle, the cel­lu­lar form of our hy­po­thet­i­cal or­gan­ism could then grow as a het­erotroph, feed­ing ghoul­ishly on the re­mains of the dead au­totrophic host cells un­til every­thing is used up. This kind of thing could per­haps oc­cur in a rather iso­lated en­vi­ron­ment, that is only rarely re-seeded by sus­cep­ti­ble host au­totrophs.
OK, now back to grad­ing un­der­grad­u­ate lab re­ports.....