GTP activator and dNTP substrates of HIV-1 restriction factor SAMHD1 generate a long-lived activated state

Abstract

The HIV-1 restriction factor sterile α-motif/histidine-aspartate domain-containing protein 1 (SAMHD1) is a tetrameric protein that catalyzes the hydrolysis of all dNTPs to the deoxynucleoside and tripolyphosphate, which effectively depletes the dNTP substrates of HIV reverse transcriptase. Here, we establish that SAMHD1 is activated by GTP binding to guanine-specific activator sites (A1) as well as coactivation by substrate dNTP binding to a distinct set of nonspecific activator sites (A2). Combined activation by GTP and dNTPs results in a long-lived tetrameric form of SAMHD1 that persists for hours, even after activating nucleotides are withdrawn from the solution. These results reveal an ordered model for assembly of SAMHD1 tetramer from its inactive monomer and dimer forms, where GTP binding to the A1 sites generates dimer and dNTP binding to the A2 and catalytic sites generates active tetramer. Thus, cellular regulation of active SAMHD1 is not determined by GTP alone but instead, the levels of all dNTPs and the generation of a persistent tetramer that is not in equilibrium with free activators. The significance of the long-lived activated state is that SAMHD1 can remain active long after dNTP pools have been reduced to a level that would lead to inactivation. This property would be important in resting CD4(+) T cells, where dNTP pools are reduced to nanomolar levels to restrict infection by HIV-1.

Keywords: dNTP induced oligomerization; enzyme catalysis; innate immunity.

Fig. 1.

Activation of dUTP hydrolysis by GTP. SAMHD1 (0.5 μM) was incubated with increasing concentrations of ³H-dUTP in the presence of five GTP activator concentrations (0.01–5 mM). (A) Time courses for dUTP hydrolysis using the C18-RP TLC plate assay. (B) Linear initial rates for formation of the ³H-dU product. (C) Initial velocities were found to follow a hyperbolic dependence on dUTP concentration with no evidence of sigmoidicity. Curves are from global nonlinear least squares best fit to all of the data using Eq. 1. (D) Secondary plot of against 1/[GTP]. (E) GTP concentration dependence of the dUTP hydrolysis rates at increasing [dUTP]. (F) Secondary plot of against [dUTP].

Fig. 2.

Self-activation, transactivation, and inhibition by dGTP. (A) dGTP self-activation of dGTP hydrolysis and transactivation of dGTP hydrolysis by 5 mM GTP. The datasets were normalized to total active sites and are indistinguishable. (B) Addition of 5 mM GTP during steady-state hydrolysis of 25 μM dGTP by 0.5 μM SAMHD1 provides no additional activation. (C) Activation and inhibition of dUTP hydrolysis by dGTP and dGTPαS. Curves were generated using the simulation program Dynafit to obtain the activation and inhibition constants for dGTP (expressions 5, 6, 7, and 8).

Fig. 3.

Dilution-jump kinetic experiments for dUTP hydrolysis. High concentrations of SAMHD1 and various nucleotide combinations were incubated as indicated in pre- and postjump reactions. (A) SAMHD1 was incubated with GTP (0.5 mM) and dUTP (5 mM), GTP alone (0.5 mM), dUTP alone (5 mM), or no nucleotide before dilution. (B) Inclusion of 0.5 mM GTP in the postjump reaction did not enhance the postjump kinetics. For comparison, the dashed line shows the curve from the dUTP-only postjump condition in A. (C) Time dependence of prejump reactions containing only dGTP (5 mM).

Fig. 4.

Dilution-jump cross-linking experiments for elucidating the oligomeric states of SAMHD1. (A) Denaturing polyacrylamide gel separation of the cross-linked species. SAMHD1 (10 μM) was incubated under the indicated prejump conditions before dilution into postjump solutions, and immediate chemical cross-linking with 50 mM glutaraldehyde was performed. Prejump (GTP = 0.5 mM; dUTP, dGTP, and dGTPαS = 5 mM) and postjump (dUTP = 1 mM) nucleotide concentrations were evaluated for their effect on oligomerization. Visualization was by silver staining, and monomer, dimer, and tetramer bands were imaged and quantified using Quantity-One. The figure is a composite of multiple gels with the order of lanes designed to clarify presentation of the data. The positions of monomer (M), dimer (D), and tetramer (T) are indicated. (B) Time dependence of dimer and tetramer stabilities under various pre- and postjump conditions. Reactions were performed as in A, and glutaraldehyde was added at the indicated times. The concentrations of pre- and postjump nucleotides were the same as stated in A.

Fig. 5.

Sedimentation velocity experiments to evaluate the effects of GTP and dGTPαS on the oligomeric state of SAMHD1. (A) SAMHD1 alone (8 μM), in the presence of 1 mM GTP or (B) SAMHD1 alone (8 μM), and in the presence of 1 mM dGTPαS were dialyzed and analyzed by sedimentation velocity centrifugation. Raw scan data were fitted to a c(S) distribution as described in Materials and Methods.

Fig. 6.

Nucleotide-dependent oligomeric equilibria of SAMHD1 and mechanism of long-lived activation. (A) Summary of the effects of GTP and dGTPαS on the oligomeric equilibria. The values for K_D and K_T are for the indicated concentration of nucleotide. (B) Model for ordered essential activation and oligomerization by GTP activators and dNTP coactivator/substrates involving the formation of a long-lived activated tetramer. The E(A1A2)₄S₄ complex is depicted as isomerizing (K_iso) to a long-lived tetramer (brackets) that performs steady-state turnover. This form only slowly reconverts to the loosely associated tetramer form that is in equilibrium with monomer, dimer, and free nucleotides. The model also indicates that the activator sites of the long-lived tetramer do not communicate with free (d)GTPs during steady-state turnover of substrate dNTPs. Although not depicted, the actively cycling tetramer does not require the presence of bound activator nucleotides (in the text).