Treating brain tumor-initiating cells using a combination of myxoma virus and rapamycin.

BACKGROUND: Intratumoral heterogeneity in glioblastoma multiforme (GBM) poses a significant barrier to therapy in certain subpopulation such as the tumor-initiating cell population, being shown to be refractory to conventional therapies. Oncolytic virotherapy has the potential to target multiple compartments within the tumor and thus circumvent some of the barriers facing conventional therapies. In this study, we investigate the oncolytic potential of myxoma virus (MYXV) alone and in combination with rapamycin in vitro and in vivo using human brain tumor-initiating cells (BTICs). METHODS: We cultured fresh GBM specimens as neurospheres and assayed their growth characteristics in vivo. We then tested the susceptibility of BTICs to MYXV infection with or without rapamycin in vitro and assessed viral biodistribution/survival in vivo in orthotopic xenografts. RESULTS: The cultured neurospheres were found to retain stem cell markers in vivo, and they closely resembled human infiltrative GBM. In this study we determined that (i) all patient-derived BTICs tested, including those resistant to temozolomide, were susceptible to MYXV replication and killing in vitro; (ii) MYXV replicated within BTICs in vivo, and intratumoral administration of MYXV significantly prolonged survival of BTIC-bearing mice; (iii) combination therapy with MYXV and rapamycin improved antitumor activity, even in mice bearing "advanced" BTIC tumors; (iv) MYXV treatment decreased expression of stem cell markers in vitro and in vivo. CONCLUSIONS: Our study suggests that MYXV in combination with rapamycin infects and kills both the BTICs and the differentiated compartments of GBM and may be an effective treatment even in TMZ-resistant patients.

Production of Myxoma virus gateway entry and expression libraries and validation of viral protein expression.

Invitrogen's Gateway technology is a recombination-based cloning method that allows for rapid transfer of numerous open reading frames (ORFs) into multiple plasmid vectors, making it useful for diverse high-throughput applications. Gateway technology has been utilized to create an ORF library for Myxoma virus (MYXV), a member of the Poxviridae family of DNA viruses. MYXV is the prototype virus for the genus Leporipoxvirus, and is pathogenic only in European rabbits. MYXV replicates exclusively in the host cell cytoplasm, and its genome encodes 171 ORFs. A number of these ORFs encode proteins that interfere with or modulate host defense mechanisms, particularly the inflammatory responses. Furthermore, MYXV is able to productively infect a variety of human cancer cell lines and is being developed as an oncolytic virus for treating human cancers. MYXV is therefore an excellent model for studying poxvirus biology, pathogenesis, and host tropism, and a good candidate for ORFeome development.

Myxoma virus: propagation, purification, quantification, and storage.

Myxoma virus (MYXV) is a member of the Poxviridae family and prototype for the genus Leporipoxvirus. It is pathogenic only for European rabbits, in which it causes the lethal disease myxomatosis, and two North American species, in which it causes a less severe disease. MYXV replicates exclusively in the cytoplasm of the host cell. Although not infectious in humans, its genome encodes proteins that can interfere with or modulate host defense mechanisms; it is able to productively infect a number of human cancer cell lines, but not normal human cells, and has also been shown to increase survival time in mouse models of human glioma. These characteristics suggest that MYXV could be a viable therapeutic agent, e.g., in anti-inflammatory or anti-immune therapy, or as an oncolytic agent. MYXV is also an excellent model for poxvirus biology, pathogenesis, and host tropism studies. It is easily propagated in a number of cell lines, including adherent cells and suspension cultures, and minimal purification is required to provide a stock for in vivo and in vitro studies.

Co-regulation of NF-kappaB and inflammasome-mediated inflammatory responses by myxoma virus pyrin domain-containing protein M013.

NF-kappaB and inflammasomes both play central roles in orchestrating anti-pathogen responses by rapidly inducing a variety of early-response cytokines and chemokines following infection. Myxoma virus (MYXV), a pathogenic poxvirus of rabbits, encodes a member of the cellular pyrin domain (PYD) superfamily, called M013. The viral M013 protein was previously shown to bind host ASC-1 protein and inhibit the cellular inflammasome complex that regulates the activation and secretion of caspase 1-regulated cytokines such as IL-1beta and IL-18. Here, we report that human THP-1 monocytic cells infected with a MYXV construct deleted for the M013L gene (vMyxM013-KO), in stark contrast to the parental MYXV, rapidly induce high levels of secreted pro-inflammatory cytokines like TNF, IL-6, and MCP-1, all of which are regulated by NF-kappaB. The induction of these NF-kappaB regulated cytokines following infection with vMyxM013-KO was also confirmed in vivo using THP-1 derived xenografts in NOD-SCID mice. vMyxM013-KO virus infection specifically induced the rapid phosphorylation of IKK and degradation of IkappaBalpha, which was followed by nuclear translocation of NF-kappaB/p65. Even in the absence of virus infection, transiently expressed M013 protein alone inhibited cellular NF-kappaB-mediated reporter gene expression and nuclear translocation of NF-kappaB/p65. Using protein/protein interaction analysis, we show that M013 protein also binds directly with cellular NF-kappaB1, suggesting a direct physical and functional linkage between NF-kappaB1 and ASC-1. We further demonstrate that inhibition of the inflammasome with a caspase-1 inhibitor did not prevent the induction of NF-kappaB regulated cytokines following infection with vMyxM013-KO virus, but did block the activation of IL-1beta. Thus, the poxviral M013 inhibitor exerts a dual immuno-subversive role in the simultaneous co-regulation of both the cellular inflammasome complex and NF-kappaB-mediated pro-inflammatory responses.

Analysis of vaccinia virus-host protein-protein interactions: validations of yeast two-hybrid screenings.

Vaccinia virus, a large double-stranded DNA virus, is the prototype of the Orthopoxvirus genus, which includes several pathogenic poxviruses of humans, such as monkeypox virus and variola virus. Here, we report a comprehensive yeast two-hybrid (Y2H) screening for the protein-protein interactions between vaccinia and human proteins. A total of 109 novel vaccinia-human protein interactions were detected among 33 viral proteins. To validate subsets of those interactions, we constructed an ORFeome library of vaccinia virus strain WR using the Gateway plasmid cloning system. By co-expressing selected vaccinia and host proteins in a variety of expression systems, we found that at least 17 of the Y2H hits identified between vaccinia and human proteins can be verified by independent methods using GST pull-down assays, representing a 63% validation rate for the Y2H hits examined (17/27). Because the cloned ORFs are conveniently transferable from the entry vectors to various destination expression vectors, the vaccinia ORFeome library will be a useful resource for future high-throughput functional proteomic experiments.

Two N-terminal regions of the Sendai virus L RNA polymerase protein participate in oligomerization.

The RNA dependent RNA polymerase of Sendai virus consists of a complex of the large (L) and phosphoprotein (P) subunits where L is thought to be responsible for all the catalytic activities necessary for viral RNA synthesis. We previously showed that the L protein forms an oligomer [Smallwood, S., Cevik, B., Moyer, S.A., 2002. Intragenic complementation and oligomerization of the L subunit of the Sendai virus RNA polymerase. Virology 304, 235-245] and mapped the L oligomerization domain between amino acids 1 and 174 of the protein [Cevik, B., Smallwood, S., Moyer, S.A., 2003. The oligomerization domain resides at the very N-terminus of the Sendai virus L RNA polymerase protein. Virology 313, 525-536]. An internal deletion encompassing amino acids 20 to 178 of the L protein lost polymerase activity but still formed an L-L oligomer. The first 25 amino acids of paramyxovirus L proteins are highly conserved and site-directed mutagenesis within this region eliminated the biological activity of the L protein but did not have any effect on P-L or L-L interactions. Moreover deletion of amino acids 2-18 in L abolished biological activity, but again the L-L binding was normal demonstrating that the oligomerization domain of L protein resides in two N-terminal regions of the protein. Therefore, sequences between both aa 2-19 and aa 20-178 can independently mediate Sendai L oligomerization, however, both are required for the activity of the protein.

Identification of a new region in the vesicular stomatitis virus L polymerase protein which is essential for mRNA cap methylation.

The vesicular stomatitis virus (VSV) L polymerase protein possesses two methyltransferase (MTase) activities, which catalyze the methylation of viral mRNA cap structures at the guanine-N7 and 2'-O-adenosine positions. To identify L sequences required for the MTase activities, we analyzed a host range (hr) and temperature-sensitive (ts) mutant of VSV, hr8, which was defective in mRNA cap methylation. Sequencing hr8 identified five amino acid substitutions, all residing in the L protein. Recombinant VSV were generated with each of the identified L mutations, and the presence of a single G1481R substitution in L, located between conserved domains V and VI, was sufficient to produce a dramatic reduction (about 90%) in overall mRNA methylation. Cap analysis showed residual guanine-N7 methylation and reduced 2'-O-adenosine methylation, identical to that of the original hr8 virus. When recombinant viruses were tested for virus growth under conditions that were permissive and nonpermissive for the hr8 mutant, the same single L mutation, G1481R, was solely responsible for both the hr and ts phenotypes. A spontaneous suppressor mutant of the rG1481R virus that restored both growth on nonpermissive cells and cap methylation was identified and mapped to a single change, L1450I, in L. Site-directed mutagenesis of the region between domains V and VI, amino acids 1419-1672 of L, followed by the rescue of recombinant viruses identified five additional virus mutants, K1468A, R1478A/D1479A, G1481A, G1481N, and G1672A, that were all hr and defective in mRNA cap methylation. Thus, in addition to the previously characterized domain VI [Grdzelishvili, V.Z., Smallwood, S., Tower, D., Hall, R.L., Hunt, D.M., Moyer, S.A., 2005. A single amino acid change in the L-polymerase protein of vesicular stomatitis virus completely abolishes viral mRNA cap methylation. J. Virol. 79, 7327-7337; Li, J., Fontaine-Rodriguez, E.C., Whelan, S.P., 2005. Amino acid residues within conserved domain VI of the vesicular stomatitis virus large polymerase protein essential for mRNA cap methyltransferase activity. J. Virol. 79, 13373-13384], a new region between L amino acids 1450-1481 was identified which is critical for mRNA cap methylation.

A single amino acid change in the L-polymerase protein of vesicular stomatitis virus completely abolishes viral mRNA cap methylation.

The vesicular stomatitis virus (VSV) RNA polymerase synthesizes viral mRNAs with 5'-cap structures methylated at the guanine-N7 and 2'-O-adenosine positions (7mGpppA(m)). Previously, our laboratory showed that a VSV host range (hr) and temperature-sensitive (ts) mutant, hr1, had a complete defect in mRNA cap methylation and that the wild-type L protein could complement the hr1 defect in vitro. Here, we sequenced the L, P, and N genes of mutant hr1 and found only two amino acid substitutions, both residing in the L-polymerase protein, which differentiate hr1 from its wild-type parent. These mutations (N505D and D1671V) were introduced separately and together into the L gene, and their effects on VSV in vitro transcription and in vivo chloramphenicol acetyltransferase minigenome replication were studied under conditions that are permissive and nonpermissive for hr1. Neither L mutation significantly affected viral RNA synthesis at 34 degrees C in permissive (BHK) and nonpermissive (HEp-2) cells, but D1671V reduced in vitro transcription and genome replication by about 50% at 40 degrees C in both cell lines. Recombinant VSV bearing each mutation were isolated, and the hr and ts phenotypes in infected cells were the result of a single D1671V substitution in the L protein. While the mutations did not significantly affect mRNA synthesis by purified viruses, 5'-cap analyses of product mRNAs clearly demonstrated that the D1671V mutation abrogated all methyltransferase activity. Sequence analysis suggests that an aspartic acid at amino acid 1671 is a critical residue within a putative conserved S-adenosyl-l-methionine-binding domain of the L protein.

The phosphoprotein (P) and L binding sites reside in the N-terminus of the L subunit of the measles virus RNA polymerase.

Measles virus encodes an RNA-dependent RNA polymerase composed of the L and P proteins. Recent studies have shown that the L proteins of both Sendai virus and parainfluenza virus 3 form an L-L complex [Cevik, B., Smallwood, S., Moyer, S.A., 2003. The oligomerization domain resides at the very Nterminus of the Sendai virus L RNA polymerase protein. Virology 313, 525-536.; Smallwood, S., Moyer, S.A., 2004. The L polymerase protein of parainfluenza virus 3 forms anoligomer and can interact with the heterologous Sendai virus L, P and C proteins. Virology 318, 439-450.; Smallwood, S., Cevik, B., Moyer, S.A., 2002. Intragenic complementation and oligomerization of the L subunit of the Sendai virus RNA polymerase. Virology 304, 235-245.]. Using differentially tagged L proteins, we show here that measles L also forms an oligomer and the L-L binding site resides in the N-terminal 408 amino acids overlapping the P binding site in the same region of L. To identify amino acids important for binding P and L, site-directed mutagenesis of the L-408 protein was performed. Seven of twelve mutants in L-408 were unable to form a complex with measles P while the remainder did bind at least some P. In contrast, all of the mutants retained the ability to form the L-L complex, so different amino acids are involved in the L and P binding sites on L. Four of the 408 mutations defective in P binding were inserted into the full-length measles L protein and all retained L-L complex formation, but did not bind P. Full-length L mutants that did not bind P were also inactive in viral RNA synthesis, showing a direct correlation between P-L complex formation and activity.

Mapping the phosphoprotein binding site on Sendai virus NP protein assembled into nucleocapsids.

To catalyze RNA synthesis, the Sendai virus P-L RNA polymerase complex first binds the viral nucleocapsid (NC) template through an interaction of the P subunit with NP assembled with the genome RNA. For replication, the polymerase utilizes an NP(0)-P complex as the substrate for the encapsidation of newly synthesized RNA which involves both NP-RNA and NP-NP interactions. Previous studies showed that the C-terminal 124 amino acids of NP (aa 401-524) contain the P-NC binding site. To further delineate the amino acids important for this interaction, C-terminal truncations and site-directed mutations in NP were characterized for their replication activity and protein-protein interactions. This C-terminal region was found in fact to be necessary for several different protein interactions. The C-terminal 492-524 aa were nonessential for the complete activity of the protein. Deletion of amino acids 472-491, however, abolished replication activity due to a specific defect in the formation of the NP(0)-P complex. Binding of the P protein of the polymerase complex to NC required aa 462-471 of NP, while self-assembly of NP into NC required aa 440-461. Site-directed mutations from aa 435 to 491 showed, however, that the charged amino acids in this region were not essential for these defects.

The L polymerase protein of parainfluenza virus 3 forms an oligomer and can interact with the heterologous Sendai virus L, P and C proteins.

We recently showed that the L protein of Sendai virus is present as an oligomer in the active P-L polymerase complex [Smallwood et al., Virology 304 (2002) 235]. We now demonstrate using two different epitope tags that the L protein of a second respirovirus, human parainfluenza type 3 virus (PIV3), also forms an L-L complex. L oligomerization requires the coexpression of the differentially epitope tagged L proteins. By exploiting a series of C-terminal truncations the L-L binding site maps to the N-terminal half of L. There is some complex formation between the heterologous PIV3 and Sendai L and P proteins; however, the heterologous L protein does not function in transcription of either the PIV3 or Sendai template. The PIV3 C protein binds PIV3 L and inhibits RNA synthesis in vitro and in vivo. Significant homology exists between the C proteins of PIV3 and Sendai and complex formation occurs between the PIV3 and Sendai heterologous C and L proteins. In addition, the heterologous C proteins can inhibit transcription at approximately 50% of the level of the homologous protein. These data suggest that while the C proteins may be functionally somewhat interchangeable, the L and P proteins are specific for each virus.

The L-L oligomerization domain resides at the very N-terminus of the sendai virus L RNA polymerase protein.

The Sendai virus RNA-dependent RNA polymerase is composed of the L and P proteins. We previously showed that the L protein gives intragenic complementation and forms an oligomer where the L-L interaction site mapped to the N-terminal half of the protein (S. Smallwood et al., 2002, Virology, 00, 000-000). We now show that L oligomerization does not depend on P protein and progressively smaller N-terminal fragments of L from amino acids (aa) 1-1146 through aa 1-174 all bind wild-type L. C-terminal truncations up to aa 424, which bind L, can complement the transcription defect in an L mutant altered at aa 379, although these L truncation mutants do not bind P. The fragment of L comprising aa 1-895, furthermore, acts as a dominant-negative mutant to inhibit transcription of wild-type L. N-terminal deletions of aa 1-189 and aa 1-734 have lost the ability to form the L-L complex as well as the L-P complex, although they still bind C protein. These data are consistent with the L-L interaction site residing in aa 1-174. Site-directed mutations in the N-terminal 347 aa, of L which abolish P binding, do not affect L-L complex formation, so while the L and P binding sites on L are overlapping they are mediated by different amino acids. The N-terminal portions of L with aa 1-424, aa 1-381, and to a lesser extent aa 1-174, can complement the transcription defect in an L mutant altered at aa 77-81, showing their L-L interaction is functional.

Intragenic complementation and oligomerization of the L subunit of the sendai virus RNA polymerase.

The RNA-dependent RNA polymerase of Sendai virus consists of two subunits, the L and P proteins, where L is thought to be responsible for all the catalytic activities necessary for viral RNA synthesis. Sequence alignment of the L proteins of a variety of negative-stranded RNA viruses revealed six regions of good conservation, designated domains I-VI, which are thought to correspond to functional domains of the protein. Analysis of a number of site-directed mutants within the six domains of L allowed us to conclude that the activities of the polymerase are not simply compartmentalized and that each domain contributes to multiple steps in viral RNA synthesis. Nevertheless these domains can function in trans since we demonstrate here that intragenic complementation between pairs of coexpressed inactive L mutants can restore viral RNA synthesis on an added template. Although intragenic complementation is typically very inefficient, complementation to restore leader RNA synthesis was surprisingly very efficient for some pairs and complementation of mRNA synthesis and genome replication was less, but still significant. Complementation occurred with L mutants in five of the six domains, the exception being a domain III mutant, and required the cotranslation of the two L mutants. C-terminal truncations deleting up to half of L were capable of restoring transcription of an inactive domain I L mutant at amino acid 379. Oligomerization of L in the polymerase complex was demonstrated directly by the co-immunoprecipitation of differentially epitope-tagged full-length and truncated L proteins. These data are consistent with L protein being an oligomer with multiple independent domains each of which exhibits several functions.

Different substitutions at conserved amino acids in domains II and III in the Sendai L RNA polymerase protein inactivate viral RNA synthesis.

The Sendai virus RNA polymerase is a complex of two virus-encoded proteins, the phosphoprotein (P) and the large (L) protein, where L is believed to possess all the enzymatic activities necessary for viral transcription and replication. The alignment of amino acid sequences of L proteins from negative-sense RNA viruses shows six regions, designated domains I-VI, of good conservation which have been proposed to be important for the various enzymatic activities of the polymerase. To directly address the role(s) of domains II and III, site-directed mutations were constructed by the substitution of multiple amino acids at 13 highly or mostly conserved residues. Analysis of in vitro viral transcription and replication showed that the majority of the mutations completely inactivated the L protein for all aspects of RNA synthesis, thus conservation correlated with the essential nature of the amino acid. At some positions different phenotypes, from inactivation to partial activities, were observed which depended on the nature of the amino acid that was substituted. Two mutants, K543R and K666V, could synthesize some leader RNA, but were defective in mRNA synthesis and replication. K666R and G737E had significantly reduced replication compared to transcription in vitro, but replicated genome RNA much more efficiently in vivo. K666A gave transcription, but no replication. Representative inactive L mutants, however, were still able to bind P protein and the polymerase complex was capable of binding nucleocapsids, so the defect appeared to be in the initiation of RNA synthesis.

The C-terminal 88 amino acids of the Sendai virus P protein have multiple functions separable by mutation.

The Sendai virus P-L polymerase complex binds the NP-encapsidated nucleocapsid (NC) template through a P-NP interaction. To identify P amino acids responsible for binding we performed site-directed mutagenesis on the C-terminal 88 amino acids in the NC binding domain. The mutant P proteins expressed from plasmids were assayed for viral RNA synthesis and for various protein-protein interactions. All the mutants formed P oligomers and bound to L protein. While two mutants, JT3 and JT8, retained all P functions at or near the levels of wild-type (wt) P, three others--JT4, JT6, and JT9--were completely defective for both transcription and genome replication in vitro. Each of the inactive mutants retained significant NC binding but had a different spectrum of other binding interactions and activities, suggesting that the NC binding domain also affects the catalytic function of the polymerase. NC binding was inhibited by combinations of the inactive mutations. The remaining P mutants were active in transcription but defective in various aspects of genome replication. Some P mutants were defective in NP(0) binding and abolished the reconstitution of replication from separate P-L and NP(0)-P complexes. In some of these cases the coexpression of the wt polymerase with the mutant NP(0)-P complex could rescue the defect in replication, suggesting an interaction between these complexes. For some P mutants replication occurred in vivo, but not in vitro, suggesting that the intact cell is providing an unknown function that cannot be reproduced in extracts of cells. Thus, the C-terminal region of P is complex and possesses multiple functions besides NC binding that can be separated by mutation.