Figure 2. HIV-1 Tat Forms Two Distinct Complexes: Tatcom1, Containing Active P-TEFb, and Tatcom2, Containing 7SK, LARP7, and MEPCE.

(A) Experimental scheme for (B) and (C).

(B) Flag-purified eTat-associated complexes were separated by centrifugation through a 15%–35% glycerol gradient. Odd numbered fractions (Fr) were analyzed by WB and assayed for in vitro CTD-kinase activity using GST-CTD and [γ-³²P]ATP. Reactions were separated on SDS-PAGE and visualized by autoradiography (bottom panel). Even-numbered fractions were probed with LARP7, AFF1, CDK9, and HA antibodies (right panel). RNA was extracted from these fractions and subjected to qRT-PCR analysis using 7SK-specific primers (top right panel). The hypophosphorylated (a) and hyperphosphorylated (o) forms of GST-CTD are indicated.

(C) Flag-purified eTat (input) was subjected to IP using AF9-, ENL-, ELL-, or LARP7-specific antibodies or irrelevant IgG. Input, IPs, as well as flowthrough (FT) from AF9 and ENL IPs were probed with the indicated antibodies (lanes 1–12). The presence of Tatcom1 subunits AFF1 and ENL in reciprocal IPs using ELL- or LARP7-specific antibodies, normalized for eTat levels, was assessed by WB (lanes 13–15).

(D) HEXIM1 and SART3 do not associate with eTat. Whole-cell extracts derived from S3 cells stably transduced with Flag- and HA-tagged LARP7, eLARP (S3LARP7), and eTat (S3Tat) were subjected to Flag IP. Cell extracts (CE) and IPs were analyzed for the presence of HEXIM1, SART3, and CDK9 by WB.

(E) Reciprocal IP with AF9 and LARP7 antibodies, normalized to amounts of CDK9 (top panel), as well as HEXIM1 IPs were assayed for in vitro kinase activity using [γ-³²P]ATP and a tetrarepeat of the CTD consensus as substrate. Reactions were separated on SDS-PAGE and visualized by autoradiography (lower panel).

(F) Schematic representation of eTat-associated complexes. See also Figure S2.