Modulation of TLR4-signaling pathway

Abstract

The present invention relates to compounds for use in modulating the toll-like receptor 4 (TLR4) signaling pathway, as well as to a pharmaceutical composition comprising said compounds.

Claims

1. A method of using a RNA oligonucleotide 2R, consisting of the nucleotide sequence shown in SEQ ID No: 2, for modulating a toll-like receptor 4 (TLR4) signaling pathway, wherein said RNA oligonucleotide down-regulates binding of heterogeneous nuclear ribonucleoprotein K (hnRNP K) to TAK1 mRNA by competition.

2. A method of administering a pharmaceutical composition, comprising a RNA oligonucleotide 2R, consisting of the nucleotide sequence shown in SEQ ID No: 2, for modulating a toll-like receptor 4 (TLR4) signaling pathway, wherein said RNA oligonucleotide down-regulates binding of heterogeneous nuclear ribonucleoprotein K (hnRNP K) to TAK1 mRNA by competition.

3. The method according to claim 2, wherein the pharmaceutical composition further comprising least one pharmaceutically acceptable solvent, diluent and/or excipient.

4. The method according to claim 1, wherein the compound is effective in the prophylactic or therapeutic treatment of sepsis.

Description

(1) The figures show:

(2) FIG. 1: Overview of the TLR4 signaling pathway

(3) Schematic representation of the TLR4 signaling pathway.

(4) FIG. 2: Time-dependent activation of macrophages by LPS

(5) Raw 264.7 cells (upper panel) and BMDM (lower panel) were treated with 10 ng/ml and 80 ng/ml LPS, respectively. mRNA synthesis of pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, IL-15, and IL-18), the anti-inflammatory cytokine IL-10 and Cox-2 as well as GAPDH or S7 was monitored by RT-PCR. Protein expression of hnRNP K and GAPDH was not affected by LPS stimulation. Cleaved caspase-3 was only detectable after 24 h, indicating that the cells did not undergo apoptosis at earlier time points.

(6) FIG. 3: Flow cytometric analysis of LPS-activated macrophages

(7) To further characterize the activation of macrophages, flow cytometric analysis of Raw 264.7 cells (left panel) and BMDM (right panel) was performed. Living cells were analyzed for the expression of specific macrophage markers (CD11b, CD115, F4/80). Positive cells showed increased expression of activation markers Gr1, MHCII, CD14, CD40, CD80 and CD86 compared to untreated cells. Annexin V staining was used to detect apoptosis.

(8) FIG. 4: mRNA translation is enhanced after LPS treatment in vitro

(9) Cytoplasmic extracts were prepared from untreated Raw 264.7 cells and after LPS stimulation. Translation activity was monitored using Luciferase (Luc) reporter constructs bearing a 5′m7G-cap and a poly(A) tail (A). An increase in Luc activity in extracts prepared from cells treated with LPS for 6 h and 10 h compared to untreated cells was observed (B). Reporter mRNAs were extracted from translation reactions and mRNA stability was measured by qRT-PCR, showing that enhanced Luc activity was not a result of increased mRNA stability (C).

(10) FIG. 5: HnRNP K interacts with c-Src and COX-2 mRNA in extracts of RAW 264.7 cells and BMDM.

(11) Upper panel: Immunoprecipitated hnRNP K from cytoplasmic extracts (CXT) of untreated RAW 264.7 cells (left) and BMDM (right) and after 6 h LPS stimulation was detected by Western blotting. For the control immunoprecipitation a non-related antibody (ctrl) was used.

(12) Lower panel: Coprecipitated mRNAs were detected by RT-PCR as indicated.

(13) FIG. 6: Verification of LPS-mediated differential interaction of specific mRNAs with hnRNP K by qRT-PCR.

(14) Equal amounts of CXT were used for immunoprecipitation with an anti-hnRNP K and a control antibody. Copurified mRNAs were analyzed by qRT-PCR. mRNAs specifically enriched on hnRNP K were normalized to five controls that did not interact with hnRNP K. Specific binding in untreated cells was compared to 6 h LPS treatment (higher values represent stronger mRNA binding in untreated cells).

(15) FIG. 7: hnRNP K sediments to mRNP fractions in extracts of untreated and LPS stimulated RAW 264.7 cells.

(16) (A) Schematic representation of the experiment. Sucrose gradient profiles of CXTs of untreated RAW 264.7 cells (B) and after 6 h LPS stimulation (C) in the presence of cycloheximide. RNA was extracted from gradient fractions and the distribution of 18S and 28S rRNA was analyzed. The distribution of rpS3, eIF6 and hnRNP K was determined by Western blotting.

(17) FIG. 8: TNFα mRNA sediments to heavy polysomes in LPS stimulated RAW 264.7 cell extracts.

(18) (A) Distribution of TNFα mRNA in sucrose gradient fractions from untreated RAW 264.7 cells and after 6 h LPS stimulation.

(19) (B) Cumulative distribution of TNFα mRNA to heavy and light polysomes, initiating ribosomes and mRNPs as indicated in (A).

(20) The compound as a well as the pharmaceutical composition of the present invention allow for the effective modulation of the TLR4-signaling pathway, particularly by modulating the binding of hnRNP K to mRNA. Such a modulation advantageously enables to specifically control the release e.g. of cytokines, without completely blocking the entire TLR4 signaling pathway. As a consequence, complications connected to severe inflammatory processes, such as sepsis, can effectively be reduced or avoided.

(21) FIG. 9: HnRNP K directly binds to the TAK1 mRNA 3′UTR.

(22) Upper panel: schematic representation of the TAK1 mRNA 3′UTR. Lower panel: 14.5 fmol [.sup.32P]-labeled TAK1 mRNA 3′UTR fragments (lanes 1-8) and the r15-LOX mRNA 3′UTR DICE (lane 9) were incubated with 7.2 pmol recombinant hnRNP K as indicated and subjected to UV-crosslinking.

(23) FIG. 10: 2R competes hnRNP K binding to TAK1 mRNA 3′UTR fragments 3 and 5.1.

(24) 40 fmol [.sup.32P]-labeled TAK1 mRNA 3′UTR fragment 3 (lanes 1-5) or 5.1 (lanes 6-10) was incubated with 7.2 pmol hnRNP K (lanes 1-10) in the presence of 10, 50, 100 or 200 fold molar excess of unlabeled competitor RNA 2R (lanes 2-5 and 7-10).

(25) FIG. 11: HnRNP K.sub.(1-334) does not interact with TAK1 mRNA 3′UTR fragments 3 and 5.1.

(26) 40 fmol [.sup.32P]-labeled TAK1 mRNA 3′UTR fragments 3 (lanes 1 and 5), 5.1 (lanes 2 and 6), r15-LOX mRNA 3′UTR DICE (lanes 3 and 7) or TAK1 mRNA 3′UTR fragment 4 (lanes 4 and 8) were incubated with 7.2 pmol recombinant His-hnRNP K.sub.wt (lanes 1-4) or His-hnRNP K.sub.(1-334) (lanes 5-8), as indicated and subjected to UV-crosslinking.

(27) FIG. 12: KH3 of hnRNP K interacts with the TAK1 mRNA 3′UTR.

(28) 40 fmol [.sup.32P]-labeled TAK1 mRNA 3′UTR fragments 3 (lane 1) and 5.1 (lane 2), r15-LOX mRNA 3′UTR DICE (lane 3) or TAK1 mRNA 3′UTR fragment 4 (lane 4) were incubated with 530 pmol recombinant KH3.

(29) FIG. 13: KH3 competes for binding of full-length hnRNP K with the r15-LOX mRNA 3′UTR DICE and TAK1 mRNA 3′UTR fragment 3.

(30) 40 fmol [.sup.32P]-labeled r15-LOX mRNA 3′UTR DICE (lanes 1-5) was incubated with 1.8 pmol full-length hnRNP K (lanes 1-4) and KH3 was added as competitor in 100, 200 and 400 fold molar excess (lanes 2-4). Lane 5 contains only the highest level of KH3 as control. 40 fmol [.sup.32P]-labeled TAK1 mRNA 3′UTR fragment 3 (lanes 6 and 7) was incubated with 7.2 pmol full-length hnRNP K (lanes 6 and 7) and KH3 was added as competitor in 74 fold molar excess (lane 7).

(31) FIG. 14: Reduction of hnRNP K results in enhanced TAK1 expression in untreated RAW 264.7 cells.

(32) A) Western blot analysis of RAW 264.7 cells, which were transfected with no siRNA (lane 1), a control siRNA (ctrl.) (lane 2) or two siRNAs directed against hnRNP K (lanes 3 and 4). Antibodies detecting hnRNP K, TAK1 and Vinculin were applied.

(33) B) Quantification of TAK1 normalized to Vinculin of the Western blot shown in A.

(34) C) qRT-PCR analysis of the mRNAs isolated from two independent transfection experiments designed as in (A) with specific primers against TAK1 mRNA. Specific target mRNA was normalized to Ndufv1 mRNA and expressed as -fold change compared to mock transfected cells, set to 1.

(35) FIG. 15: Reduction of hnRNP K does not affect TAK1 mRNA stability in untreated RAW 264.7 cells.

(36) Analysis of TAK1 mRNA stability by qRT-PCR. Cells were transfected with control siRNA (si-ctrl., triangles, dashed line) or hnRNP K siRNA (267, squares, solid line), treated with actinomycin D for 0, 1 and 2 h. TAK1 mRNA was normalized to Ndufv1 and GAPDH mRNAs (0 h time point 100%).

(37) FIG. 16: Reduction of hnRNP K enhances TAK1 mRNA translation in untreated RAW 264.7 cells.

(38) Top panel: Western blot analysis of cytoplasmic extracts generated from untreated RAW 264.7 cells, which were transfected with a control siRNA (ctrl.) or hnRNP K siRNA (267) with antibodies specific for hnRNP K and Vinculin.

(39) Middle panel: A representative A.sub.260 nm profile of cytoplasmic extracts characterized in the upper panel fractionated on 15-45% sucrose density gradients is shown. Polysomes, 80S ribosomes, 60S and 40S ribosomal subunits and mRNPs are indicated. RNA extracted from the 19 gradient fractions was visualized by gelred staining of an agarose gel. 18S and 28S rRNA are indicated. The distribution of rpS3 and eIF6 in the sucrose gradient fractions was determined by Western blotting.

(40) Bottom panel: Endogenous TAK1 mRNA distribution in cytoplasmic extract generated from cells transfected with ctrl. (triangles, dashed line) or hnRNP K (267) siRNA (squares, solid line) was determined by qRT-PCR using the ΔCt-method and normalized to exogenously added LUC mRNA extraction control. The percentage of TAK1 mRNA in each fraction is shown.

(41) FIG. 17: LPS treatment of macrophages with reduced hnRNP K level that led to enhanced TAK1 protein expression shows an earlier and prolonged p38 phosphorylation and a stronger cytokine mRNA synthesis.

(42) A) Western blot analysis of lysates generated from untreated RAW 264.7 cells, which were transfected with a control siRNA (ctrl.) (lanes 1-8) or hnRNP K siRNA (267) (lanes 9-16) with antibodies specific for hnRNP K, TAK1, ERK1/2, phospho-ERK1/2, p38, phospho-p38 and Vinculin.

(43) B) Endogenous TNF-α, IL-1β and IL-10 mRNAs were determined by qRT-PCR using the ΔΔCt-method and normalized to Ndufv1 mRNA.

(44) FIG. 18: Schematic representation of AKT2 mRNA 3′UTR.

(45) FIG. 19: 2R competes binding of hnRNP K to the Akt2 mRNA 3′UTR.

(46) 40 fmol [.sup.32P]-labeled Akt2 mRNA fragment 1 (lanes 1-5), Akt2 mRNA 3′UTR fragment 2 (lanes 6-10), r15-LOX mRNA 3′UTR DICE (lane 11) or TAK1 mRNA 3′UTR fragment 4 (lane 12) was incubated with 7.2 pmol hnRNP K (lanes 1-12) in the presence of 10, 50, 100 or 200 fold molar excess of unlabeled competitor RNA 2R (lanes 2-5 and lanes 7-10).

(47) FIG. 20: HnRNP K.sub.(1-334) does not interact with the Akt2 mRNA 3′UTR.

(48) 40 fmol [.sup.32P]-labeled fragments of the Akt2 mRNA 3′UTR as indicated (lanes 1, 2 and lanes 5, 6), the DICE of the r15-LOX mRNA (lanes 3 and 7) or TAK1 mRNA 3′UTR fragment 4 (lanes 4 and 8) was incubated with 3.6 pmol recombinant hnRNP K.sub.wt (lanes 1-4) or hnRNP K.sub.(1-334) (lanes 5-8). (*) marks an unspecific band.

(49) FIG. 21: KH3 of hnRNP K interacts with the Akt2 mRNA 3′UTR.

(50) 40 fmol [.sup.32P]-labeled fragments of the Akt2 mRNA 3′UTR as indicated (lanes 1 and 2), the DICE of the r15-LOX mRNA (lane 3) and the TAK1 mRNA 3′UTR fragment 4 (lane 4) were incubated with 530 pmol recombinant KH3.

(51) FIG. 22: Reduction of hnRNP K does not influence Akt2 mRNA level or mRNA stability.

(52) A) Western blot analysis of untreated RAW 264.7 cells (lanes 1-4) or after 6 h LPS treatment (lanes 5-8), which were transfected with no siRNA (lanes 1 and 5), a control siRNA (ctrl.) (lanes 2 and 6) or two siRNAs directed against hnRNP K (lanes 3, 4, 7 and 8) with antibodies against hnRNP K and Vinculin.

(53) B) qRT-PCR analysis of the mRNAs isolated from the transfection experiments designed as in A) with specific primers against Akt2 mRNA. Specific target mRNA was normalized to Ndufv1 mRNA and expressed as -fold change compared to mock transfected cells, set to 1.

(54) C) Analysis of Akt2 mRNA stability by qRT-PCR. Cells were transfected with control siRNA (si-ctrl., triangles, 0h black dashed line, 6 h LPS grey dashed line) or hnRNP K siRNA (267, squares, 0 h black solid line, 6 h LPS grey solid line), treated with actinomycin D for 0, 1 and 2 h. Akt2 mRNA was normalized to Ndufv1 mRNA (0 h time point 100%).

(55) FIG. 23: Reduction of hnRNP K reduces Akt2 mRNA translation in untreated RAW 264.7 cells and after 6 h LPS.

(56) A) Top panel: Western blot analysis of cytoplasmic extracts generated from untreated RAW 264.7 cells or after 6 h LPS treatment, which were transfected with a control siRNA (ctrl.) or hnRNP K siRNA (267) with antibodies specific for hnRNP K and Vinculin. Middle panel: Representative A260 nm profiles of cytoplasmic extracts characterized in the upper panel fractionated on 15-45% sucrose density gradients is shown. Polysomes, 80S ribosomes, 60S and 40S ribosomal subunits and mRNPs are indicated. RNA extracted from the 19 gradient fractions was visualized by gelred staining of an agarose gel. 18S and 28S rRNA are indicated. The distribution of rpS3 and eIF6 in the sucrose gradient fractions was determined by Western blotting.

(57) B) Endogenous Akt2 mRNA distribution in pooled gradient fractions as indicated was determined by qRT-PCR using the ΔCt-method and normalized to exogenously added LUC mRNA extraction control. The percentage of Akt2 mRNA is shown.

(58) C) Endogenous Ndufv1 mRNA distribution in pooled gradient fractions as indicated was determined by qRT-PCR using the ΔCt-method and normalized to exogenously added LUC mRNA extraction control. The percentage of Ndufv1 mRNA is shown.

(59) FIG. 24: LPS treatment of cells with reduced hnRNP K level that led to reduced Akt2 protein expression showed a stronger Arg1 mRNA synthesis.

(60) A) Western blot analysis of lysates generated from untreated RAW 264.7 cells, which were transfected with a control siRNA (ctrl.) (lanes 1-8) or hnRNP K siRNA (267) (lanes 9-16) with antibodies specific for hnRNP K and Vinculin.

(61) B) Endogenous Arg1 mRNA was determined by qRT-PCR using the ΔΔCt-method and normalized to Ndufv1 mRNA.

(62) The present invention will now be further illustrated in the following examples without being limited thereto.

EXAMPLES

Example 1

(63) To obtain insight into the impact of post-transcriptional control of gene expression on LPS induced TLR4 signaling, in particular the function of regulatory mRNA protein complexes, RAW 264.7 cell culture was employed, which allowed the generation of the number of cells that is necessary to produce extracts for biochemical experiments. Additionally, protocols were established to isolate and cultivate BMDM from healthy C57BL/6 mice as a tool to validate newly identified mRNAs and regulatory proteins involved in the post-transcriptional control of TLR4 signaling. It was decided to use C57BL/6 mice, because several relevant knock-out models were derived from this strain.

(64) The required LPS concentration to induce RAW 264.7 cells (10 ng/ml) and BMDM (80 ng/ml) was determined in titration experiments. When RAW 264.7 cells were treated with LPS, the time dependent induction of ERK and p38 could be detected. Interestingly, BMDM from healthy C57BL/6 mice showed an earlier response than RAW 264.7 cells. Furthermore, a time dependent induction of mRNAs that encode pro-inflammatory cytokines (TNFα, IL-1β, IL-6, IL-15, IL-18), the anti-inflammatory cytokine IL 10 and COX-2 mRNA was observed, as monitored by RT-PCR. The expression of hnRNP K was not affected by LPS stimulation in both cell types, as confirmed by Western blotting. The caspase-3 cleavage product occurred only after 24 hrs, indicating that the cells did not undergo apoptosis at earlier time points (FIG. 2). For further analyses, time points 0, 20 min, 1, 2, 6 and 10 h were chosen.

(65) To characterize activated macrophages, FACS analysis was performed. Living cells were analyzed for the expression of CD11b, CD115 and F4/80 (specific macrophage markers) with a FACSCanto II. Positive cells showed increased expression of activation markers (Gr1, MHCII, CD14, CD40, CD80 and CD86) compared to untreated cells (FIG. 3).

(66) Further, cytoplasmic extracts were prepared from untreated RAW 264.7 cells and after LPS stimulation. Translation activity was monitored using Luciferase (Luc) reporter constructs bearing a 5′m7G-cap and a poly(A) tail. An increase in Luc activity in extracts prepared from cells treated with LPS for 6 h and 10 h compared to untreated cells was observed. Reporter mRNAs were extracted from translation reactions and mRNA stability was measured by qRT-PCR, showing that enhanced Luc activity was not a result of increased mRNA stability (FIG. 4).

(67) To identify new translational regulated mRNAs by microarray analysis, it was focused on hnRNP K as a regulatory RNA binding protein. hnRNP K was immunoprecipitated with an anti-hnRNP K antibody from cytoplasmic extracts (CXT) of untreated RAW 264.7 cells and BMDM or from cells stimulated with LPS for 6 h, but not with a nonrelated control antibody (FIG. 5, upper panel). Endogenous c-Src mRNA coprecipitated with hnRNP K, as shown before in K562 cells. Interestingly, c-Src mRNA is not present in untreated RAW 264.7 cells and BMDM, but its synthesis is induced by LPS. Previously in THP-1 cells, it has been shown that hnRNP K interacts with the COX-2 mRNA that was used as a second positive control. Myosin X served as negative control.

Example 2

(68) After immunoprecipitation, mRNAs were purified, converted to cDNAs and analyzed on an Affymetrix Mouse Genome 430 2.0 and 3′IVT array. By comparing input, anti-hnRNP K and control immunoprecipitation of untreated RAW 264.7 cells or after 6 h LPS stimulation, each in duplicates, 1901 differentially regulated mRNAs were found. Interestingly, 177 mRNAs were identified that encode proteins involved in immune response.

(69) To further validate the list of newly identified hnRNP K-associated mRNAs, 21 candidate mRNAs (Tab. 1) were initially selected that encode factors acting mostly in the TLR4 signaling and associated downstream pathways and their expression was analyzed by qRT-PCR.

(70) TABLE-US-00001 TABLE 1 Identified target mRNAs that were specifically enriched in hnRNP K immunoprecipitation. # Gene Symbol Gene Title Gene ID 1 Pcgf1, Nspc1, polycomb group ring 69837 AU024121 finger 1 2 Pcgf2, Mel18, polycomb group ring 22658 Rnf110, Zfp144 finger 2 3 Med9, Med25 mediator of RNA Poly- 192191 merase II transcription, subunit 9 homolog (yeast) 4 Irak1, IRAK, Plpk, interleukin-1 receptor- 16179 mPLK, Il1rak associated kinase 1 5 Carm1, Prmt4 coactivator-associated 59035 arginine methyl- transferase 1 6 Litaf, TBX1, LPS-induced TN factor 56722 N4WBP3 7 Icam2, CD102, intercellular adhesion 15896 Ly-60 molecule 2 8 Fcer1g, Fce1g, Fc receptor, IgE, high 14127 Ly-50 affinity I, gamma polypeptide 9 Alox5, 5LO, arachidonate 5- 11689 5LX, 5-LOX lipoxygenase 10 CD115, Csf1r, colony stimulating 12978 Fms, Csfmr, factor 1 receptor Fim-2, M-CSFR 11 Pik3ca, p110, phosphatidylinositol 18706 caPI3K, p110alpha 3-kinase, catalytic, alpha polypeptide 12 Irak1bp1, Aabp3, interleukin-1 receptor- 65099 Aip70, Simpl associated kinase 1 binding protein 1 13 Irak4, NY-REN-64 interleukin-1 receptor- 266632 associated kinase 4 14 Tak1, Map3k7 mitogen-activated 26409 protein kinase kinase kinase 7 15 Nkrf NF-kappaB repressing 77286 factor 16 Erc1, ELKS, ELKS/RAB6-interacting/ 111173 Rab6ip2 CAST family member 1 17 Jun, AP-1, Junc, Jun oncogene 16476 c-jun 18 Akt1, Akt, PKB, thymoma viral proto- 11651 Rac, PKB/Akt, oncogene 1 PKBalpha 19 Akt2, PKB, thymoma viral proto- 11652 PKBbeta, Akt2 oncogene 2 20 Akt3 hymoma viral proto- 23797 oncogene 3 21 IRF7 interferon regulatory 54123 factor 7

Example 3

(71) As a first criterion of exclusion, it was tested whether the selected 21 mRNAs show an equal expression in untreated RAW 264.7 cells and after 6 h LPS stimulation, indicating that transcriptional control of these mRNAs on the one hand and regulation of mRNA translation on the other can be distinguished. For 13 mRNAs, an equal amplification could be detected from the RNA pool isolated from control and LPS-treated RAW 264.7 cells, which was used for hnRNP K immunoprecipitation (data not shown).

(72) Next, it was investigated whether the 13 equally expressed mRNAs show a specific enrichment with hnRNP K in untreated cells, but not LPS stimulated cells. Five mRNAs, for which no interaction with hnRNP K could be detected in the microarray analysis, served as negative controls (data not shown). For seven mRNAs with equal expression levels, a stronger enrichment on hnRNP K was determined in untreated RAW 264.7 cells compared to 6 h LPS stimulation. In contrast, the other five mRNAs were bound stronger by hnRNP K after LPS stimulation, while one mRNA was equally bound (FIG. 6).

(73) Among the first group of seven, mRNAs that encode IRAK4, TAK1, IRAK1BP1, Erc1, CARM1/PRMT4, PIK3ca and Akt3 were identified. The kinases IRAK4 and TAK1 play essential roles in the MyD88 dependent TLR4 signaling pathway. IRAK1BP1 has an inhibitory function in inflammation. Erc1 is an IκB kinase regulatory subunit that is required for NFκB activation and CARM1/PRMT4 has been shown to function as a promoter-specific regulator of NFκB. LPS is implicated in the activation of the PI3K and Akt pathway in monocytes/macrophages. The PI3K family consists of three classes: IA, B, II and III. A dependence of LPS induced Akt activation on class IA PI3K has been shown by RNAi mediated PIK3ca silencing in monocytes.

Example 4

(74) Next it was hypothesized that silencing of target mRNA translation by hnRNP K is abolished when the protein is released upon LPS stimulation. To address this issue, investigation was started as to whether LPS stimulation affects the differential distribution of target mRNAs to polysomes and translational inactive mRNPs through the use of sucrose gradient analysis of RAW 264.7 CXTs (FIG. 7).

(75) Before harvesting, untreated or 6 h LPS stimulated RAW 264.7 cells were treated with cycloheximide, which inhibits the peptidyltransferase reaction and stalls translating ribosomes. CXTs prepared from these cells were fractionated on 15-45% sucrose gradients (FIG. 7A).

(76) The distribution of 18S and 28S rRNA was used to analyze the position of ribosomal complexes and mRNPs (FIG. 7B, C). 28S and 18S rRNA accumulated in bottom fractions 2 to 8 (polysomes) and in fractions 13 and 14 (80S ribosomes), while fractions 16 and 17 contained mainly 18S rRNA of 40S ribosomal subunits (FIG. 7B, C). EIF6, a translation initiation factor that is bound to the ribosomal 60S subunit prior to 80S ribosome formation, was enriched in fraction 15. HnRNP K was mainly distributed to fractions 18 and 19, indicating that it is associated with mRNPs in untreated RAW 264.7 cells and after 6 h LPS stimulation (FIG. 7B, C). The enhanced distribution of ribosomal RNAs to polysomal fractions after 6 h of LPS treatment indicates an increase in mRNA translation (FIG. 7C).

Example 5

(77) As a proof of principle, the distribution of the TNFα mRNA in the sucrose gradient profiles was analyzed next. Consistent with previous results, a shift of the TNFα mRNA from light polysomes to heavy polysomes was observed after 6 h LPS stimulation (FIG. 8).

Example 6

(78) For two mRNAs that encode the critical kinases TAK1 and Akt2, which act in different branches of the TLR4 signaling pathway, the experimental strategy is summarized herein.

(79) With the experiments outlined below it is shown that: hnRNP K interacts directly with specific sequence elements in the 3′ untranslated region (3′UTR) of TAK1 and Akt2 mRNA. Binding of hnRNP K modulates TAK1 mRNA and Akt2 mRNA translation specifically in an LPS-dependent manner. Translation control by hnRNP K results in enhanced TAK1 and reduced Akt2 synthesis, which is in agreement with their differential functions in TLR4 downstream signaling. LPS-induced modulation of TAK1 and Akt2 mRNA translation by hnRNP K has a downstream impact on the pro- and anti-inflammatory cytokine expression.

(80) Importantly, the hnRNP K mRNA interaction can be competed with: a small peptide (79aa) a small nucleic acid (38nts), as shown in FIGS. 10, 13 and 19.

(81) The analysis of mRNAs that are equally expressed in untreated and LPS-activated macrophages revealed that binding of hnRNP K to 13 mRNAs that encode mostly TLR4 downstream signaling proteins is significantly enhanced or decreased (FIG. 6). These mRNAs are potential targets of hnRNP K-mediated translational regulation and changes in the respective protein levels will affect TLR4 downstream-signaling, which results in modulated pro- and anti-inflammatory cytokine expression.

(82) Two mRNAs were chosen to prove that the synthesis of TLR4 downstream signaling proteins is regulated by hnRNP K in an LPS-dependent manner and the resulting modulation of TLR4 signaling affects pro- and anti-inflammatory cytokine expression: Tak1: Transforming growth factor-β (TGFβ) activated kinase 1 (TAK1) mRNA was stronger enriched on hnRNP K in untreated RAW 264.7 cells compared to 6 h LPS stimulation (cf. FIG. 6). Akt2: In contrast, a stronger binding to hnRNP K was determined for Akt2 mRNA after LPS stimulation (cf. FIG. 6).
TAK1

(83) TAK1, a MAP3K member, is the central kinase of the MyD88-dependent branch in TLR4 signaling (FIG. 1), which leads to the activation of IκB kinases (IKK) in the NFκB pathway (early response) and the three MAPK pathways: ERK 1/2, c-Jun N-terminal kinase (INK) and p38. MyD88 recruits the kinases IRAK1 and IRAK4. IRAK1, a target of IRAK4 recruits tumor necrosis factor (TNF) receptor associated factor 6 (TRAF6) to the TLR4-receptor complex. The IRAK1-TRAF6 complex dissociates from the TLR4-receptor complex and activates TAK1. TAK1 phosphorylates MKK4, MKK3/6 and IKK, which activate JNK and p38 pathways and induce IκB degradation leading to NFκB activation. These signaling pathways upregulate transcription factors including NFκB and AP-1, implicated in the coordinated induction of pro-inflammatory (TNFα, IL-1β) and anti-inflammatory (IL-10) cytokine encoding genes.

(84) TAK1 mRNA 3′ untranslated region (UTR) was screened for the presence of potential hnRNP K binding sequence motifs. It contains several CU-rich (U/CCCC) elements, which were first identified as hnRNP K binding motifs in the differentiation control element (DICE) (241nts) of reticulocyte 15-Lipoxygenase (r15-LOX) mRNA 3′UTR. HnRNP K is a potent regulator of r15-LOX mRNA translation in erythroid cell maturation. DICE binding is conferred by the third hnRNP K homology domain (KH3) of hnRNP K. The TAK1 mRNA 3′UTR (>3600 nts) were divided into eight fragments (1.1, 1.2, 2, 3, 4, 5.1, 5.3 and 6) to analyze direct interaction with hnRNP K in in vitro UV-crosslinking assays with [.sup.32P]-labeled transcripts (FIG. 9, lanes 1-8). Fragments 3, 5.1 and 5.3 were strongly bound by recombinant His-hnRNP K, the DICE served as positive control (FIG. 9, lanes 4, 6, 7 and 9). Interestingly, fragment 4 that contains a class I AU-rich element (ARE) did not interact (FIG. 9, lane 5).

(85) Signals for fragment 5.1 and 5.3 were comparable and fragment 6 did not interact, suggesting that U/CCCC elements in fragment 5.1 mediate the interaction with hnRNP K. Next, it was investigated whether a short fragment of the DICE (2R) (38nts) that is sufficient to mediate r15-LOX mRNA translation regulation is able to compete hnRNP K binding to TAK1 mRNA fragments 3 and 5.1. 2R did compete the interaction of hnRNP K with fragments 3 and 5.1 (FIG. 10, lanes 2-5 and 7-10), further indicating that hnRNP K binds to the CU-rich sequence.

(86) Previously, it has been shown that DICE-binding of hnRNP K lacking KH3 (hnRNP K.sub.(1-334)) was significantly reduced. To prove that deletion of KH3 results in loss of interaction with TAK1 mRNA fragments 3 and 5.1, UV-crosslinking of [.sup.32P]-labeled fragments 3 and 5.1 with full length His-hnRNP K and truncated His-hnRNP K.sub.(1-334) was performed. Fragment 4 served as negative and the DICE as positive control (FIG. 11). HishnRNP K was specifically bound to fragment 3 and 5.1 and the DICE, but not fragment 4 (FIG. 11, lanes 1-4). In contrast, binding of His-hnRNP K.sub.(1-334) to fragment 3, 5.1 and the DICE RNA was strongly reduced (FIG. 11, lanes 5-8). Residual binding to the DICE results from non-specific interaction of KH1 and KH2 in the absence of KH3.

(87) In order to prove that KH3 directly interacts with TAK1 mRNA fragments 3 and 5.1, His-KH3 was employed. Again, the DICE served as positive, fragment 4 as negative control (FIG. 12). His-KH3 specifically binds to fragment 3 and 5.1 and the DICE, but not fragment 4 (FIG. 12, lanes 1-4).

(88) Next, it was investigated whether KH3 (79 amino acids) could function as a small peptide competitor that interferes with binding of full-length hnRNP K (463 amino acids). For this purpose the DICE or fragment 3 was exposed to full-length hnRNP K and recombinant KH3 in the same reaction (FIG. 13). KH3 competes for binding of full-length hnRNP K to the DICE or TAK1 mRNA 3′UTR fragment 3 (FIG. 13, lanes 1-7).

(89) Having shown that recombinant hnRNP K directly interacts with the TAK1 mRNA 3′UTR, it was next investigated whether hnRNP K acts as a cellular regulator of TAK1 mRNA translation in murine macrophages (RAW 264.7). For this purpose, RNAi with two individual siRNAs that target hnRNP K (267, 1467) was performed. As a control, a nonrelated siRNA (ctrl.) was used (FIG. 14). Transfection of non-treated RAW 264.7 cells with the specific siRNAs strongly enhanced the expression of TAK1 protein (FIGS. 14 A and B).

(90) To address the mechanism underlying enhanced TAK1 protein expression at reduced endogenous hnRNP K levels, it was first examined whether hnRNP K depletion resulted in an increase of TAK1 mRNA (FIG. 14C). TAK1 mRNA levels were not affected, indicating that enhanced transcription of the TAK1 gene did not account for elevated TAK1 protein synthesis (FIGS. 14 A and B). The stability of TAK1 mRNA could also be affected by hnRNP K.

(91) To assess TAK1 mRNA stability after siRNA-mediated hnRNP K reduction, transcription was blocked by actinomycin D treatment. TAK1 mRNA stability remained unchanged when hnRNP K siRNA (267) or a control (ctrl.) siRNA were applied (FIG. 15). Taken together, these observations suggest that the increase in TAK1 protein is a consequence of enhanced Tak1 mRNA translation when hnRNP K is reduced.

(92) To prove that endogenous TAK1 mRNA translation is up-regulated in cells transfected with the hnRNP K siRNA (267), TAK1 mRNA co-sedimentation with polysomes was characterized by fractionation of cytoplasmic extracts on 15 to 45% sucrose gradients (FIG. 16, bottom panel). MRNA translation was stalled by cycloheximide treatment before RAW 264.7 cells were harvested for extract preparation. The efficiency of the hnRNP K knock down in RAW 264.7 cells was analyzed by Western blotting (FIG. 16, top panel).

(93) The distribution of 18S and 28S rRNA and Western blot analysis of rpS3, a ribosomal protein of the 40S ribosomal subunit and translation initiation factor eIF6 that is bound to the ribosomal 60S subunit prior to 80S ribosome formation were used to analyze the position of ribosomal complexes and mRNPs (FIG. 16, middle panel). 28S and 18S rRNA accumulated in bottom fractions 2 to 8 (polysomes) and in fractions 13 and 14 (80S ribosomes), while fractions 16 and 17 contained mainly 18S rRNA of 40S ribosomal subunits. RpS3 accumulated in 40S and 80S and complex-containing fractions (FIG. 16, middle panel). EIF6 was enriched in fraction 15 that contains 60S ribosomal subunits (FIG. 16, middle panel). The analysis of endogenous TAK1 mRNA distribution by qRT-PCR illustrates that hnRNP K knock down resulted in a strong accumulation of TAK1 mRNA in polysomal fractions (FIG. 16, bottom panel), indicating an enhanced translation of TAK1 mRNA in vivo. The results clearly show that hnRNP K down regulation enhances endogenous TAK1 mRNA translation resulting in an increase of TAK1 protein.

(94) To obtain insight in the function of the regulatory mRNA protein complex it was investigated whether the reduction of hnRNP K that leads to enhanced TAK1 protein expression results in increased phosphorylation of downstream MAP kinases (p38 or ERK1/2) and to elevated cytokine mRNA synthesis. For this purpose, RAW 264.7 cells, which were transfected with a nonrelated control siRNA (ctrl.) (FIG. 17 A, left panel) or with the specific siRNA against hnRNP K (267) (FIG. 17 A, right panel) were treated with 10 ng/ml LPS for the indicated time points. A time dependent induction of ERK and p38 phosphorylation in RAW 264.7 cells, which were transfected with a control siRNA (ctrl.) (FIG. 17 A, left panel) was observed. Interestingly, knock down of hnRNP K by the specific siRNA (267) led to an earlier and prolonged phosphorylation of p38, but did not affect phosphorylation of ERK1/2 (FIG. 17 A, compare left and right panel). Importantly, synthesis of TNF-α, IL-1β and IL-10 mRNAs increased after reduction of hnRNP K by RNAi (FIG. 17 B).

(95) Taken together, the above results indicate that binding of hnRNP K to the TAK1 mRNA 3′UTR inhibits TAK1 mRNA translation in untreated RAW 264.7 cells. The reduction of endogenous hnRNP K leads to an increase in TAK1 mRNA translation. When cells in which TAK1 protein expression is enhanced are treated with LPS an earlier and prolonged phosphorylation of p38 and an increase of cytokine mRNA synthesis can be detected, suggesting that hnRNP K is an important modulator of TAK1 mRNA translation that affects TLR-4 downstream signaling and induction of cytokine mRNA expression.

(96) Akt2

(97) LPS induced TLR-4 signaling activates the phosphatidylinositol 3 kinase (PI3K)/Akt pathway. The first kinase recruited and activated after PI3K activation is 3-phosphoinositide-dependent kinase (PDK-1). PDK-1 interacts and phosphorylates members of the Akt family of serine/threonine protein kinases (Akt1, Akt2 and Akt3). Activation of Akt (also known as PKB) results in the phosphorylation of a number of substrates that have potential importance in LPS signaling, like glycogen synthase kinase (GSK-3), caspase 9, IκB kinases. The three Akt family members are closely related and highly conserved. Recently it has been shown that specific down regulation of Akt2 level impaired chemotaxis of both THP-1 and mouse peritoneal macrophages, suggesting that Akt2 is required for macrophage trafficking. Furthermore it was reported that Akt kinases differentially contribute to macrophage polarization, with Akt1 knockout giving rise to an M1 and Akt2 ablation resulting in an M2 phenotype.

(98) As shown before for TAK1 mRNA, the 3′UTR of the Akt2 mRNA was screened for the presence of potential hnRNP K binding elements. It contains several U/CCCC motifs (FIG. 18).

(99) To analyze whether recombinant hnRNP K directly interacts with the Akt2 mRNA 3′UTR, UV-crosslinking experiments were performed with the [.sup.32P]-labeled Akt2 mRNA 3′UTR subdivided in two fragments (1 and 2) (FIG. 19, lanes 1-5 and 6-10). Recombinant hnRNP K was bound strongly to both fragments (FIG. 19, lanes 1 and 6) and to the DICE, but not to the TAK1 mRNA 3′UTR fragment 4 that served as negative control (FIG. 19, lanes 11 and 12). Furthermore, it was investigated whether the short RNA consisting of 2 repeats of the DICE (2R) can be used to compete the binding of hnRNP K to the Akt2 mRNA 3′UTR fragments 1 and 2. As shown in FIG. 19, the hnRNP K-Akt2 mRNA interaction could be competed by the short ribonucleic acid 2R (FIG. 19, lanes 2-5 and 7-10), indicating that hnRNP K binds to the CU-rich sequence.

(100) KH3 of hnRNP K is required for specific RNA interaction and RNA binding of hnRNP K lacking KH3 (hnRNP K.sub.(1-334)) is strongly reduced. To prove that the KH3 peptide also mediates the interaction with Akt2 mRNA 3′UTR fragments 1 and 2, in vitro RNA binding was employed. His-hnRNP Kwt and the deletion variant His-hnRNP K.sub.(1-334) were incubated with [.sup.32P]-labeled Akt2 mRNA 3′UTR fragments 1 and 2. The DICE served as positive and TAK1 mRNA 3′UTR fragment 4 as negative control (FIG. 20). Whereas His-hnRNP K.sub.wt is bound to the Akt2 mRNA 3′UTR fragment 1, 2 and DICE RNA, but not to TAK1 mRNA 3′UTR fragment 4 (FIG. 20, lanes 1-4), binding of His-hnRNP K.sub.(1-334) to fragment 1, 2 and DICE RNA is strongly reduced (FIG. 20, lanes 5-8).

(101) To prove that the peptide KH3 specifically and directly interacts with Akt2 mRNA fragment 1 and 2, His-KH3 was incubated with [.sup.32P]-labeled Akt2 mRNA fragment 1 and 2. The DICE and TAK1 mRNA 3′UTR fragment 4 were employed as positive and negative controls, respectively (FIG. 21, lanes 1-4). The peptide KH3 specifically interacts with Akt2 mRNA fragment 1 and 2 and DICE RNA, but not the TAK1 mRNA fragment (FIG. 21, lanes 1-4).

(102) Having shown that recombinant hnRNP K directly interacts with the Akt2 mRNA 3′UTR, it was next investigated whether endogenous hnRNP K acts as a cellular regulator of Akt2 mRNA translation in murine RAW 264.7 macrophages in an LPS dependent manner. For this purpose, RNAi with two siRNAs that target hnRNP K (267, 1467), and a non-related control siRNA (ctrl.) were performed. The efficiency of the hnRNP K knock down was analyzed by Western blotting (FIG. 22A).

(103) Next, it was examined whether hnRNP K depletion resulted in a reduced Akt2 mRNA level or decreased Akt2 mRNA stability. Akt2 mRNA levels and stability remained unchanged after hnRNP K knock down (FIGS. 22 B and C). To determine whether polysomal loading of Akt2 mRNA is affected by hnRNP K, Akt2 mRNA co-sedimentation was characterized with polysomes, monosomes or mRNPs in sucrose gradient analysis (FIG. 23).

(104) RAW 264.7 cells, transfected with a nonrelated control siRNA (ctrl) or hnRNP K siRNA (267) were left untreated or treated for 6 h with LPS. MRNA translation was stalled by cycloheximide treatment before the cells were harvested. Cytoplasmic extracts were fractionated on 15 to 45% sucrose gradients. The efficiency of the hnRNP K knock down in RAW 264.7 cells is shown in the Western blot inserts (FIG. 23 A, upper panel). Fractions that contain polysomes (1-12), monosomes (12-17) and mRNPs (18-19) were pooled (FIG. 23 A, lower panel) to analyze endogenous Akt2 mRNA distribution. QRT-PCR analysis illustrates that hnRNP K knock down resulted in a strong accumulation of Akt2 mRNA in fractions containing monosomes and mRNPs (FIG. 23 B), indicating a decreased translation of Akt2 mRNA in untreated RAW 264.7 cells (left panel) that is further enhanced after 6 h LPS treatment (right panel). In contrast, translation of Ndufv1 mRNA remained unchanged (FIG. 23 C). These studies indicate that siRNA-mediated down regulation of endogenous hnRNP K results in an inhibition of endogenous Akt2 mRNA translation. Akt2 could not be detected immunologically, because a suitable antibody was not available.

(105) To determine macrophage-specific functional consequences of hnRNP K mediated Akt2 regulation it was investigated whether reduced Akt2 mRNA translation caused by hnRNP K down regulation results in an M2 macrophage phenotype. M2 macrophage polarization and expression of arginase 1 (Arg1) are primarily regulated by the transcription factors C/EBPβ and STATE. C/EBPβ levels are elevated in Akt2.sup.−/− macrophages or RAW 264.7 cells in which Akt2 expression is silenced compared with wildtype or Akt1-deficient cells. Increased C/EBPβ binding to the Arg1 promoter leads to an induction of Arg1 mRNA synthesis in Akt2.sup.−/− macrophages. RAW 264.7 cells, which were transfected with a nonrelated control siRNA (ctrl.) (FIG. 24 A, left panel) or with the specific siRNAs against hnRNP K (267) (FIG. 24 A, right panel) were treated with 10 ng/ml LPS for the indicated time points.

(106) We could observe a time dependent induction of Arg1 mRNA synthesis in the control cells, which is significantly elevated in hnRNP K knock down cells (FIG. 24 B). Taken together, these results indicate that hnRNP K is an important modulator of Akt2 mRNA translation.

(107) Discussion:

(108) Pathogen components, such as bacterial LPS, that act through TLR4 lead to the activation of MAPKs and NFκB through different pathway branches to subsequently induce pro- and anti-inflammatory cytokine expression. The expression of kinases and their modulators involved in these processes is controlled both transcriptionally and post-transcriptionally. Post-transcriptional regulation of gene expression is mediated by trans-acting factors, like RNA-binding proteins at the level of mRNA stability and mRNA translation. This facilitates rapid responses to environmental changes.

(109) To obtain insight into the function of regulatory mRNA-protein complexes RAW 264.7 cells and bone marrow derived macrophages (BMDM) from C57BL/6 mice were used. Macrophage maturation after LPS stimulation was confirmed by flow cytometry (FACS) and cytokine mRNA levels by RT-PCR.

(110) To identify mRNAs, which are translationally regulated in an LPS dependent manner, cytoplasmic extracts (CXT) from untreated RAW 264.7 cells and BMDM and after 6 h LPS stimulation were generated. From these extracts a known modulator of mRNA translation, heterogeneous nuclear ribonucleoprotein (hnRNP) K was immunoprecipitated with a specific antibody. Coimmunoprecipitated mRNAs were applied to microarray analysis (Affymetrix Mouse Genome 430 2.0 and 3′IVT array). 1901 mRNAs were identified, which are differentially enriched on hnRNP K upon LPS treatment. Interestingly, 177 mRNAs encode proteins involved in the immune response. Initially, 21 candidate mRNAs were selected that encode mostly factors acting in the TLR4 pathway and their expression was analyzed by qRT-PCR. 13 mRNAs were equally amplified from the RNA pool isolated from control and LPS-treated RAW 264.7 cells, suggesting that these are regulated post-transcriptionally. A stronger specific enrichment on hnRNP K was determined in untreated RAW 264.7 cells compared to LPS stimulation for seven of these mRNAs, whereas five mRNAs were bound stronger by hnRNP K after LPS treatment. The first group of identified candidate target mRNAs encode IL-1R associated kinase-4 (IRAK4), transforming growth factor-β (TGFβ) activated kinase 1 (TAK1), IRAK1-binding protein 1 (IRAK1BP1), IκB kinase regulatory subunit (Erc1/ELKS), coactivator-associated arginine methyltransferase 1 (CARM1/PRMT4), phosphatidylinositol 3 kinase (PI3K) catalytic subunit p110a (PIK3ca) and thymoma viral protooncogene 3 (Akt3). It is therefore apparent that hnRNP K silences target mRNA translation in untreated macrophages, while the protein is released from the mRNA upon LPS stimulation. Earlier it has been shown that hnRNP K is a specific activator and substrate of the tyrosine kinase c-Src. c-Src-dependent phosphorylation of Y458 in hnRNP K leads to loss of specific mRNA binding and abolishes its function as an inhibitor of mRNA translation.