Peripheral expression of Jak3 is required to maintain T lymphocyte function.

The Jak family tyrosine kinase, Jak3, is involved in signaling through cytokine receptors that utilize the common gamma chain (gammac), such as those for IL-2, IL-4, IL-7, IL-9, and IL-15. Recent studies of Jak3-deficient mice and humans have demonstrated that Jak3 plays a critical role in B and T lymphocyte maturation and function. The T lymphocyte defects in Jak3-deficient mice include a small thymus, a decrease in peripheral CD8+ cells, an increase in the surface expression of activation markers, and a severe reduction in proliferative and cytokine secretion responses to mitogenic stimuli. To determine whether the peripheral T lymphocyte defects result from aberrant maturation in the thymus or from the absence of Jak3 protein in peripheral T cells, we generated reconstituted mice that express normal levels of Jak3 protein in the thymus but lose Jak3 expression in peripheral T cells. Jak3 expression in the thymus restores normal T cell development, including CD8+, gammadelta, and natural killer cells. However, the loss of Jak3 protein in peripheral T cells leads to the Jak3-/- phenotype, demonstrating that Jak3 is constitutively required to maintain T cell function.

Defects in B lymphocyte maturation and T lymphocyte activation in mice lacking Jak3.

Biochemical studies of signaling mediated by many cytokine and growth factor receptors have implicated members of the Jak family of tyrosine kinases in these pathways. Specifically, Jak3 has been shown to be associated with the interleukin-2 (IL-2) receptor gamma chain, a component of the receptors for IL-2, IL-4, IL-7, IL-9, and IL-15. Mice lacking Jak3 showed a severe block in B cell development at the pre-B stage in the bone marrow. In contrast, although the thymuses of these mice were small, T cell maturation progressed relatively normally. In response to mitogenic signals, peripheral T cells in Jak3-deficient mice did not proliferate and secreted small amounts of IL-2. These data demonstrate that Jak3 is critical for the progression of B cell development in the bone marrow and for the functional competence of mature T cells.

The role of Jak3 in lymphoid development, activation, and signaling.

Mutations in a number of lymphoid signaling molecules lead to immunodeficiencies in mice and humans. Among these, one very pleiotropic syndrome results from deficiencies in an array of cytokine signaling pathways utilizing a cytokine receptor common gamma chain, gammac, and the tyrosine kinase Jak3. Recent advances in our understanding of the role of gammac and Jak3 in lymphocyte development and function highlight the importance of cytokine receptor signaling pathways in regulating lymphoid homeostasis and responsiveness.

T cell development and activation in Jak3-deficient mice.

Jak3, a member of the Janus family of tyrosine kinases, participates in signaling through cytokine receptors that contain the common gamma-chain, including the receptors for interleukin (IL)-2, IL-4, IL-7, IL-9, and IL-15. Jak3- and gamma c-deficient mice have pleiotropic defects that can be attributed to their inability to respond to multiple specific cytokines. A great deal of recent work has focused on the T cell defects in these mutant mice. Specifically, Jak3- and gamma c-deficient mice have small thymuses revealing a defect in T cell development, and in addition, have functionally unresponsive peripheral T cells with an activated/memory cell phenotype. The thymic defect in these mutant mice strongly resembles that seen in IL-7 and IL-7 receptor knockout mice, suggesting that the lack of IL-7 receptor signaling accounts for this defect in Jak3-/- and gamma c- mice. To characterize this defect further, we have examined the earliest stages of T cell development in adult and fetal Jak3-/- thymuses. These studies identify two discrete developmental defects at the CD4-CD8- stage of T cell maturation. Analyses of peripheral T cells in Jak3-/- and gamma c- mice have also revealed a number of abnormalities. All of the T cells in these mutant mice have an activated phenotype and a large fraction of them are proliferating in vivo. In addition, Jak3-/- and gamma c- T cells are more prone to undergo apoptosis than wild-type T cells. Together, these features account for the decreased IL-2 secretion by in vitro-stimulated Jak3-/- T cells. Overall, many of the lymphoid defects of Jak3- and gamma c-deficient mice can be accounted for by the lack of IL-7R and IL-2R signaling; however, other cytokine systems must also be involved in maintaining peripheral T cell homeostasis.

Genomic structure and promoter region of the murine Janus-family tyrosine kinase, Jak3.

Genomic DNA sequences encoding the murine Janus family tyrosine kinase Jak3 were isolated to determine the intron-exon structure of the gene and to investigate the phylogeny of Jak-family kinases. The murine Jak3 gene comprises approximately 15 kbp of genomic DNA and consists of 23 exons. The organization of sequences encoding the pseudo-kinase domain of Jak3 is similar to the intron-exon structure encoding catalytic domains of Src-family tyrosine kinases, whereas the pattern of introns-exons encoding the Jak3 kinase domain shows no structural similarity to that of other tyrosine kinase genes. Genomic analysis further indicates that alternative splicing gives rise to different forms of the murine Jak3 mRNA encoding different isoforms of the Jak3 protein. Analysis of Jak3 intron-exon structure also suggests that a mutation in the human JAK3 gene responsible for a severe combined immune deficiency (SCID) phenotype results from aberrant splicing of the JAK3 transcript. Finally, potential regulatory sequences in the upstream region of the murine Jak3 gene were analyzed and are discussed in relation to the known expression pattern of Jak3.

Mechanism of interferon action. Translational control and the RNA-dependent protein kinase (PKR): antagonists of PKR enhance the translational activity of mRNAs that include a 161 nucleotide region from reovirus S1 mRNA.

The interferon-inducible, RNA-dependent protein kinase (PKR) is an important regulator of viral protein synthesis. Activated PKR inhibits protein synthesis by phosphorylating initiation factor eIF-2 alpha. The reovirus S4 gene, whose 1196 nucleotide mRNA transcript does not activate the PKR kinase, is efficiently expressed in vector-transfected monkey COS cells. By contrast, the 1463 nucleotide S1 gene of reovirus, which is a potent activator of PKR, is poorly expressed in COS cells. Virus genetic engineering was therefore used to examine the effect of the PKR activator sequence from the reovirus S1 gene on the expression of chimeric genes of reovirus in transfected COS cells. Chimeric S1/S4 and S4/S1/S4 reovirus constructions that included the PKR activator sequence from S1 in the sigma 3 ORF of S4 were expressed much less efficiently than wild-type S4. However, expression of sigma 3 from S4 (3'UTR/S1), which included the PKR activator sequence from S1 within the 3'-UTR of S4, was comparable to that from wild-type S4. Treatment of COS cells with 2-aminopurine, an inhibitor of PKR, increased the expression of the reovirus S1, S1/S4, and S4/S1/S4 chimeric genes but not the S4 gene or S4 (3'UTR/S1) chimera in transfected COS cells. Likewise, coexpression of the phosphotransfer-negative mutant PKR (K296R) increased the expression of reovirus S1, S1/S4 and S4/S1/S4 chimeric genes but not the S4 gene or S4 (3'UTR/S1) chimera in cotransfected COS cells. Truncated PKR(1-243) which includes the dsRNA binding domain but not the kinase catalytic subdomains was able to enhance the expression of reovirus S1, but did not affect S4 expression. The dsRNA binding protein E3L encoded by vaccinia virus also increased S1 expression similar to PKR (1-243) and PKR(K296R). These results suggest that the translational repression in vivo mediated by PKR is selective for mRNAs that possess the kinase activator region, and that the dominant negative effect of PKR on gene expression is likely mediated by the RNA binding activity of the PKR protein.

Mechanism of interferon action: evidence for intermolecular autophosphorylation and autoactivation of the interferon-induced, RNA-dependent protein kinase PKR.

The interferon-induced RNA-dependent protein kinase (PKR) is postulated to have an important regulatory role in the synthesis of viral and cellular proteins. Activation of the enzyme requires the presence of a suitable activator RNA and is accompanied by an autophosphorylation of PKR. Active PKR phosphorylates the alpha subunit of protein synthesis eukaryotic initiation factor 2, resulting in an inhibition of translation initiation. The mechanism of autophosphorylation is not well understood. Here we present evidence that the autophosphorylation of human PKR can involve intermolecular phosphorylation events, i.e., one PKR protein molecule phosphorylating a second PKR molecule. Both wild-type PKR and the point mutant PKR(K296R) synthesized in vitro were phosphorylated, even though PKR(K296R) was deficient in kinase catalytic activity. Phosphorylation of both wild-type PKR and PKR(K296R) was inhibited in the presence of 2-aminopurine. Furthermore, purified human recombinant PKR(K296R) was a substrate for the purified wild-type human PKR kinase. This intermolecular phosphorylation of mutant PKR(K296R) by wild-type PKR was dependent on double-stranded RNA and was inhibited by 2-aminopurine. Finally, PKR mRNA was capable of mediating an autoactivation of wild-type PKR kinase autophosphorylation in vitro.

Mechanism of interferon action: autoregulation of RNA-dependent P1/eIF-2 alpha protein kinase (PKR) expression in transfected mammalian cells.

The expression of a molecular cDNA clone (P1 KIN) of the human RNA-dependent P1/eIF-2 alpha protein kinase (PKR) was examined in transfected monkey cells and in cell-free protein-synthesizing systems. Expression of the wild-type (wt) P1 KIN cDNA, which encodes an active protein kinase, was compared with that of the phosphotransfer catalytic domain II Lys-296-->Arg (K296R) mutant cDNA, which does not encode an active kinase. wt and K296R mutant P1 mRNAs prepared by transcription in vitro with T7 RNA polymerase programmed the cell-free synthesis of P1 ribosome-associated protein with comparable efficiency in the rabbit reticulocyte system. The K296R mutant P1 protein was also efficiently synthesized in vivo in transfected COS monkey cells. However, synthesis of the wt P1 protein was reduced about 30-fold in transfected COS cells as compared with the K296R mutant P1 protein. Cotransfection of wt P1 KIN cDNA with either K296R mutant P1 KIN cDNA or reovirus S4 cDNA greatly reduced the synthesis of K296R mutant P1 protein and reovirus sigma 3 protein, respectively. Although the wt and K296R mutant P1 KIN plasmid expression vectors replicated with comparable efficiencies in COS cells, the steady-state amount of P1 mRNA was about 3-fold less in COS cells transfected with the wt as compared with the K296R mutant P1 KIN cDNA. These results suggest that RNA-dependent P1 protein kinase expression is autoregulated in vivo in transfected mammalian cells primarily at the level of translation by a mechanism that is likely dependent upon catalytically active P1 kinase.

Mechanism of interferon action: characterization of the intermolecular autophosphorylation of PKR, the interferon-inducible, RNA-dependent protein kinase.

The interferon-inducible, RNA-dependent protein kinase (PKR) is activated by autophosphorylation, a process mediated by double-stranded RNA. A catalytically deficient, histidine-tagged mutant PKR protein [His-PKR(K296R)] was used as the substrate for characterization of the intermolecular phosphorylation catalyzed by purified wild-type PKR [PKR(Wt)]. The intermolecular autophosphorylation of His-PKR(K296R) by PKR(Wt) was RNA dependent. Excess His-PKR(K296R) substrate inhibited both the auto- and the trans-phosphorylation activities of PKR(Wt). Inhibition of PKR(Wt) by His-PKR(K296R) was relieved by higher concentrations of activator double-stranded RNA. Phosphopeptide analysis revealed that the sites of intermolecular autophosphorylation in His-PKR(K296R) were very similar, if not identical, to the sites that were autophosphorylated in PKR(Wt) and suggest a multiple of four major phosphorylation sites per PKR molecule.

The Jak family tyrosine kinase Jak3 is required for IL-2 synthesis by naive/resting CD4+ T cells.

The Jak family tyrosine kinase, Jak3, is involved in signaling through cytokine receptors using the common gamma-chain. Mice deficient in Jak3 have mature T cells, all of which have an activated/memory cell phenotype but are unresponsive to in vitro stimulation. Due to this activated phenotype, it has been impossible to determine whether Jak3 plays a role in the responsiveness of naive/resting T cells. To circumvent this difficulty, we generated naive/resting Jak3-negative T cells by two genetic approaches. After stimulation, these cells failed to produce significant amounts of IL-2. Although no signaling defect could be detected, we did find that naive/resting Jak3-negative T cells have substantially reduced levels of the transcription factor NF-AT1 and moderately reduced levels of c-Jun and c-Fos. On the basis of these data, we propose that Jak3-dependent cytokine signals may be required to maintain the normal levels of basal transcription factors required for immediate responsiveness to Ag activation.

Isolation and characterization of a molecular cDNA clone of a human mRNA from interferon-treated cells encoding nucleolar protein B23, numatrin.

A cDNA clone encoding human nucleolar phosphoprotein B23, numatrin, was isolated from a library prepared with mRNA from human U cells. The complete nucleotide sequence was determined; it revealed a single open reading frame of 294 amino acids that included four in-frame AUG codons as potential sites of translation initiation. Comparison of the human B23 nucleotide sequence with the rat and mouse B23 sequences revealed 91% homology (hum:rat, and hum:mus) in the coding region; the predicted B23 proteins displayed 94% amino acid identity. Northern gel blot analysis revealed a single B23 mRNA species of approximately 1.5 kb. The level of B23 mRNA in U cells was not detectably altered by treatment with either alpha or gamma interferon. Southern gel blot analysis revealed polymorphism within the human B23 gene structure, and suggested the presence of multiple B23 genes and/or extensive splicing of B23 RNA transcripts.

T cells from Jak3-deficient mice have intact TCR signaling, but increased apoptosis.

The Jak family tyrosine kinase, Jak3, is involved in signaling through cytokine receptors utilizing the common gamma-chain (gamma(c)). Mice and humans lacking Jak3 have severe immune deficiencies, including defects in B and T lymphocyte development and function. In particular, Jak3-deficient mice have mature T cells with an activated phenotype, yet these cells are functionally nonresponsive. In this work, we show that Jak3-deficient T cells have no gross defects in early T cell signaling, as measured by TCR-induced tyrosine phosphorylation and calcium mobilization responses. Furthermore, we find that Jak3-deficient mice expressing a transgenic TCR have extremely low numbers of peripheral T cells with a naive phenotype, indicating that both peripheral activation and expansion of Jak3-/- T cells are driven by antigenic signals. We show that, similar to gamma(c)-deficient mice, T cells from Jak3-deficient mice have an increased susceptibility to apoptosis. Previously, we showed that when stimulated in vitro by CD3 plus CD28 cross-linking, Jak3-/- T cells secrete greatly reduced amounts of IL-2, and fail to proliferate. However, by measuring intracellular IL-2 levels, we find that Jak3-/- T cells produce amounts of IL-2 similar to activated T cells from control mice, further supporting the notion that there is no defect in TCR signaling in Jak3-/- T cells.

Mechanism of interferon action: cDNA structure, expression, and regulation of the interferon-induced, RNA-dependent P1/eIF-2 alpha protein kinase from human cells.

A molecular cDNA clone (P1 KIN) was isolated that encodes the human RNA-dependent P1/eIF-2 alpha protein kinase. The complete cDNA sequence of the P1 KIN cDNA was determined; the longest open reading frame (ORF) encoded a 551 amino acid protein with a deduced molecular weight of 62055 Da. Transcripts prepared from the P1 KIN cDNA by transcription in vitro with T7 RNA polymerase programmed the cell-free synthesis of a protein indistinguishable by immunoprecipitation and immunoblot gel analyses from the authentic 67-kDa P1 protein synthesized in human U cells treated with interferon (IFN). Furthermore, by use of a sensitive primer extension assay with T7 DNA polymerase, the major site of translation initiation within the deduced ORF of the P1 KIN cDNA was directly identified. Northern RNA gel-blot analysis revealed that the P1 KIN cDNA strongly hybridized to two IFN-induced mRNAs present in both human amnion U cells and HeLa cells; their sizes were 2.5 and 6 kb. Both transcripts were efficiently induced by IFN-alpha, but poorly by IFN-gamma. Polyclonal antibody was prepared against the product of the P1 KIN cDNA expressed in Escherichia coli. In Western blot analysis the antibody recognized a 67-kDa protein induced in human cells by IFN-alpha and, in addition, a 90-kDa protein whose level was not greatly altered by IFN treatment. The IFN-induced 67-kDa protein was found associated with the ribosomal salt-wash fraction of IFN-treated human cells, whereas the 90-kDa protein was predominantly in the S100 soluble fraction. The time course for the induction by IFN-alpha of RNA-dependent protein P1 kinase activity measured by immunoprecipitation was comparable to the time course for protein P1 induction measured by Western immunoblot analysis. The amino acid sequence of P1/eIF-2 alpha protein kinase deduced from the cDNA was 62% identical with the 518-residue murine TIK kinase and contained, within the carboxy-terminal half of the protein, the motifs commonly conserved among protein-serine/threonine kinases. The amino-terminal half of the P1 protein did not possess conserved kinase motifs, but did show extensive homology with vaccinia virus-predicted protein E3L.

Mechanism of interferon action: identification of a RNA binding domain within the N-terminal region of the human RNA-dependent P1/eIF-2 alpha protein kinase.

A molecular cDNA clone of the human RNA-dependent P1/eIF-2 alpha protein kinase was expressed in Escherichia coli. Mutant P1 proteins were examined for RNA binding activity by Northwestern blot analysis using the reovirus s1 mRNA, an activator of the kinase; the adenovirus VAI RNA, an inhibitor of kinase activation; or human immunodeficiency virus (HIV) TAR RNA as probe. Analysis of TrpE-P1 deletion mutant fusion proteins revealed that the 11-kDa N-terminal region of the P1 protein bound reovirus s1 mRNA, adenovirus VAI RNA, and HIV TAR RNA. Neither s1 RNA, VAI RNA, nor TAR RNA was bound by truncated P1 proteins which lacked the N-terminal 98 amino acids. Computer analysis revealed that the human protein P1 sequence corresponding to amino acid residues within the N-terminal RNA binding domain displays high homology (greater than 54% identity; 61 to 94% similarity) with two animal virus proteins which possess RNA binding activity (vaccinia virus E3L; rotavirus VP2) and two proteins of unknown function (murine TIK; rotavirus NS34), but which are likely RNA binding proteins.

Mechanism of interferon action. cDNA structure and regulation of a novel splice-site variant of the catalytic subunit of human protein kinase A from interferon-treated human cells.

A molecular cDNA clone (KIN27) was isolated that encodes a novel variant of the catalytic subunit C alpha of the human cyclic AMP-dependent protein kinase (PKA). Analysis of PKA genomic sequence data revealed that the KIN27 variant transcripts likely result from alternative splice-site selection of the C alpha gene transcripts. The KIN27 cDNA has therefore been designated C alpha 2. RNA amplification by polymerase chain reaction revealed that KIN27 C alpha 2 splice-site variant RNA and PKA C alpha RNA were both expressed in the amnion U and HeLa human cell lines. C alpha RNA was about 3- to 5-fold more abundant than C alpha 2 RNA. Interferon treatment decreased the steady-state amount of the C alpha 2 RNA relative to C alpha RNA. These results suggest that alternative splicing may contribute to the structural heterogeneity of C subunits expressed in human cells and that interferon may affect this process.

Mechanism of interferon action: double-stranded RNA-specific adenosine deaminase from human cells is inducible by alpha and gamma interferons.

Treatment of human amnion U cells with interferon increased the steady state level of mRNA encoding the double-stranded (ds) RNA-specific adenosine deaminase (AdD) as measured by Northern gel-blot analysis. A single major dsRNA-specific AdD transcript of approximately 6.7 kb was detected; the transcript was induced by both interferon-alpha (IFN-alpha) and interferon-gamma (IFN-gamma). Likewise, Western immunoblot analysis revealed that a 150-kDa protein recognized by antiserum prepared against recombinant dsRNA-specific AdD was increased in the human amnion U and neuroblastoma SH-SY5Y cell lines treated with interferon. Both IFN-alpha and IFN-gamma induced the 150-kDa protein. These results, which establish that dsRNA-specific AdD is an IFN-inducible protein in human cells, have implications regarding the possible role of interferon in persistent viral infections.

A Fas-based suicide switch in human T cells for the treatment of graft-versus-host disease.

Graft-versus-host disease (GVHD) is a major complication of allogeneic bone marrow transplantation. One strategy to treat GVHD is to equip donor T cells with a conditional suicide mechanism that can be triggered when GVHD occurs. The herpes simplex virus thymidine kinase (HSV-tk)/ganciclovir system used clinically has several limitations, including immunogenicity and cell cycle dependence. An alternative switch based on chemically inducible apoptosis was designed and evaluated. A chimeric human protein was expressed comprising an extracellular marker (DeltaLNGFR), the Fas intracellular domain, and 2 copies of an FK506-binding protein (FKBP). Primary human T lymphocytes retrovirally transduced with this construct could be purified to homogeneity using immunomagnetic beads. Genetic integrity of the construct was ensured by redesigning repetitive sequences. Transduced T cells behaved indistinguishably from untransduced cells, retaining the ability to mount a specific antiallogeneic immune response. However, they rapidly underwent apoptosis with the addition of subnanomolar concentrations of AP1903, a bivalent "dimerizer" drug that binds FKBP and induces Fas cross-linking. A single 2-hour treatment eliminated approximately 80% of T cells, and multiple exposures induced further apoptosis. T cells were eliminated regardless of their proliferation state, suggesting that the AP1903/Fas system, which contains only human components, is a promising alternative to HSV-tk for treating GVHD.

Mechanism of interferon action. Biochemical and genetic evidence for the intermolecular association of the RNA-dependent protein kinase PKR from human cells.

The interferon-inducible protein kinase (PKR) is activated by an RNA-dependent autophosphorylation. Structure-function studies of the 551 amino acid PKR kinase from human cells have revealed that catalytic-deficient PKR mutants such as PKR(1-551)K296R display a dominant negative behavior when expressed in transfected cells. The potential for PKR to form protein multimers has therefore been examined. Three types of studies, including both genetic and biochemical analyses, demonstrated that PKR from human cells undergoes an intermolecular association that is not dependent upon RNA. First, the intermolecular association of PKR in vitro was demonstrated in the context of an enzyme-substrate interaction. Purified recombinant histidine-tagged PKR(1-551)K296R mutant protein was phosphorylated by purified wild-type PKR; this intermolecular phosphorylation of PKR was dependent on double-stranded RNA. At a fixed RNA concentration, high concentrations of the HIS-PKR(1-551)K296R mutant both impaired the autophosphorylation of wild-type PKR and blocked the trans-phosphorylation of itself. Second, the yeast two-hybrid system was used to probe the intermolecular association of PKR in vivo. Coexpression of the full-length catalytic-deficient phosphotransfer mutant PKR(1-551)K296R as a fusion protein with the Gal4 activation domain and the Gal4 DNA binding domain resulted in the expression of two Gal4-responsive reporter genes, HIS3 and lacZ. The full-length RNA-binding deficient PKR(1-551)K64E/K296R double mutant also interacted with PKR(1-551)K296R sufficiently to activate Gal4-responsive reporter genes; however, other PKR mutants including PKR(1-280)wt and PKR(281-551)K296R as well as p53, RAS, and BCL2 did not. Third, both PKR(1-551)K296R and PKR(1-551)K64E/K296R enhanced the expression of the reovirus S1 gene and S1/S4 chimeric gene in cotransfected COS cells. By contrast, the expression of the reovirus S4 gene was not enhanced by cotransfection with either PKR(1-551)K296R or PKR(1-551)K64E/K296R. These results indicate that PKR interacts with itself in an intermolecular manner both in vivo and in vitro, and that RNA binding is neither necessary nor sufficient for PKR multimerization.

Characterization of the heparin-mediated activation of PKR, the interferon-inducible RNA-dependent protein kinase.

Heparin can substitute for double-stranded (ds) RNA in the autophosphorylation and activation of the interferon-inducible, RNA-dependent elF-2 alpha protein kinase (PKR). We have used heparin oligosaccharides of defined lengths to examine the heparin-mediated activation of human PKR. Heparin oligosaccharide with 8 sugar residues was nearly as efficient as 16-residue heparin (Hep-16) in mediating the activation of PKR autophosphorylation, whereas 6-residue heparin was a poor activator. When examined in combination, Hep-16 and dsRNA did not act synergistically in activating PKR autophosphorylation. The RNA-binding activity of recombinant PKR, measured with adenovirus VA RNA, was competed by poly(rl):poly(rC) but not by Hep-16. When the catalytically inactive, histidine-tagged mutant PKR protein [His-PKR(K296R)] was examined as a substrate for purified wild-type PKR, the intermolecular phosphorylation of His-PKR(K296R) was efficiently catalyzed by dsRNA-activated PKR but not by heparin-activated PKR. However, elF-2 alpha phosphorylation was catalyzed by both heparin-and dsRNA-activated PKR. Preincubation of PKR with Hep-16 in the absence of ATP blocked subsequent autophosphorylation mediated either by Hep-16 or dsRNA, whereas preincubation with dsRNA either alone or in combination with Hep-16 did not impair subsequent autophosphorylation. Neither Hep-16 nor dsRNA caused a detectable degradation of PKR during preincubation or subsequent autophosphorylation of PKR. These results suggest that, while both dsRNA and heparin are capable of activating PKR autophosphorylation, the structural and functional basis of PKR activation differs for these two classes of polyanionic biomolecules.