Targeted modified TNF family members

Abstract

The present invention relates to a modified cytokine of the TNF superfamily, with reduced activity to its receptor, wherein said modified cytokine is specifically delivered to target cells. Preferably, said modified cytokine is a single chain variant of the TNF superfamily, even more preferably, one or more of the chains can-y one or more mutations, resulting in a low affinity to the receptor, wherein said mutant cytokine is specifically delivered to target cells. The targeting is realized by fusion of the modified cytokine of the TNF superfamily to a targeting moiety, preferably an antibody or antibody-like molecule. The invention relates further to the use of such targeted modified cytokine of the TNF superfamily to treat diseases.

Claims

1. A composition comprising a proteinaceous construct, comprising: (i) a single chain polypeptide comprising three modified human TNFs, wherein: each modified human TNF comprises a modified amino acid residue by substitution at Y115 position relative to wild type human TNF (SEQ ID NO: 14), the substitution being selected from A or G; and the modified human TNFs have reduced affinity towards their receptor as compared to wild type human TNF; and (ii) a targeting moiety that is an antibody or a variable domain of a camelid heavy chain antibody (VHH) directed to a neo-vasculature tissue or cancer tissue specific marker, wherein the composition has significant biological activity towards cells that are targeted by the targeting moiety.

2. The composition of claim 1, wherein the targeting moiety is a VHH.

3. The composition of claim 1, wherein the targeting moiety is directed towards CD20.

4. The composition of claim 3, wherein the targeting moiety is a VHH.

5. The composition of claim 4, wherein the Y115 substitution is Y115A.

6. The composition of claim 4, wherein the Y115 substitution is Y115G.

7. The composition of claim 1, wherein the targeting moiety is directed towards Her2.

8. The composition of claim 7, wherein the targeting moiety is a VHH.

9. The composition of claim 8, wherein the Y115 substitution is Y115A.

10. The composition of claim 8, wherein the Y115 substitution is Y115G.

11. The composition of claim 1, wherein the Y115 substitution is Y115A.

12. The composition of claim 1, wherein the Y115 substitution is Y115G.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1: Representation of the structural elements of the different set-ups of the sc hTNF-nanobody fusion protein.

(2) FIG. 2: Firefly luciferase activity induced by the indicated sc hTNF preparations, as compared to WT hTNF, on HekT cells (panel A) or Hek-mLR cells (panel B). Both were transiently transfected with the NF-KB luciferase reporter.

(3) FIG. 3: Firefly luciferase activity induced by the indicated sc hTNF preparations carrying a linker with 6 GGS repeats on HekT cells (panel A) or Hek-mLR cells (panel B). Both were transiently transfected with the NF-KB luciferase reporter.

(4) FIG. 4: Firefly luciferase activity induced by the indicated sc hTNF preparations carrying a linker with 13 GGS repeats on HekT cells (panel A) or Hek-mLR cells (panel B). Both were transiently transfected with the NF-KB luciferase reporter.

(5) FIG. 5: Firefly luciferase activity induced by the indicated sc hTNF preparations carrying a linker with 19 GGS repeats on HekT cells (panel A) or Hek-mLR cells (panel B). Both were transiently transfected with the NF-KB luciferase reporter.

(6) FIG. 6: Fold induction of IL-6 mRNA levels upon treatment of SK-BR-3 cells with 500 ng/ml of the indicated sc hTNF preparations compared to the levels in untreated cells and cells stimulated with the Her2 nanobody. Data represents the mean±SD of 2 independent experiments (n=4).

(7) FIG. 7: % activity of hTNF mutants compared to WT hTNF as measured by toxicity on MCF7 cells (a breast cancer cell line).

(8) FIG. 8: Toxicity on MCF7 (panel A) or MCF7-mLR (panel B) cells of targeted modified TNFs coupled to mLR NB (NB C-terminally of TNF).

(9) FIG. 9 Toxicity on MCF7 (panel A) or MCF7-hCD20 (panel B) cells of targeted modified TNFs coupled to hCD20 NB (NB C-terminally of TNF).

(10) FIG. 10: Toxicity on MCF7 (panel A) or MCF7-hCD20 (panel B) cells of targeted modified TNFs coupled to hCD20 or control BclI10 NB (NB N-terminally of TNF).

(11) FIG. 11: Toxicity on MCF7 (panel A) or MCF7-hCD20 (panel B) cells of targeted modified TNFs containing sc hTNF with combined mutations (NB N-terminally of TNF).

(12) FIG. 12: Toxicity on MCF7 (panel A) or MCF7-mLR (panel B) cells of targeted modified TNF with individual trimerizing chains coupled to mLR NB (NB C-terminally of TNF).

(13) FIG. 13: Toxicity on MCF7 (panel A) or MCF7-hCD20 (panel B) cells of targeted modified TNF with individual trimerizing chains coupled to hCD20 NB (NB N-terminally of TNF).

(14) FIG. 14: Comparison of the in vivo toxicity of WT hTNF versus targeted WT and modified sc hTNFs coupled to hCD20 or control BclI10 NB (NB N-terminally of TNF). (A) Hypothermia (B) Mortality.

(15) FIG. 15: In vivo toxicity of WT or modified (Y86F3x) sc mouse (m)TNF coupled to control BclI10 NB (NB N-terminally of TNF). (A) Hypothermia (B) Mortality.

(16) FIG. 16: In vivo anti-tumor effect of WT or modified (Y86F3x) sc mouse (m)TNF coupled to mCD20 or control BclI10 NB (NB N-terminally of TNF). (A) Tumor growth (B) Mortality.

EXAMPLES

(17) Materials and Methods to the Examples

(18) Nanobodies

(19) The nanobody 4-10 directed against the murine leptin receptor (mLR) was described in Zabeau et al. (2012). Its coding sequence is cloned into the mammalian expression vector pMET7 (Takebe et al., 1988) in fusion with the SIgK leader peptide, the HA tag and albumin. Plasmid name: pMET7 S1gK-HA-4.11-Albumin. The anti-Her2 nanobody 1R59B was described in Vaneycken et al. (2011). The NB 2HCD25 directed against the human CD20 (hCD20) and the 2MC57 NB against mouse CD20 (mCD20) were generated using standard techniques (Gharouhdi et al., 1997; Pardon et al., 2014). The control NB BclI10 was described in De Groeve et al. (2010).

(20) scTNF

(21) scTNF that consists of three hTNF monomers coupled via GGGGS-linkers (SEQ ID NO: 1) has been described by Boschert et al. (2010). The Y87Q mutation in hTNF was shown to completely abrogate the binding to both receptors, TNF-R1 and TNF-R2. Mutating 197 results in reduced binding of hTNF to both receptors (Loetscher et al., 1993). A whole range of residues within hTNF were mutated (QuikChange Site-Directed Mutagenesis Kit, Stratagene Cat #200518) and tested for their toxic effects on MCF7 cells (FIG. 7). We selected the following mutations for the targeted constructs: Y87Q, Y87F, 197S, 197A, Y115A, Y115G. The coding sequences of sc hTNF WT-6xGGS, sc hTNF Y87Q3x-6xGGS, sc hTNF I97A3x-6xGGS, sc hTNF Y87Q1x I97A2x-6xGGS, sc hTNF Y87Q2x I97A1x-6xGGS, sc hTNF WT1 x Y87Q2x-6xGGS, sc hTNF WT2x Y87Q1x-6xGGS, sc hTNF 19753x-6xGGS, sc hTNF Y115A3x-6xGGS, sc hTNF Y87F3x, sc hTNF Y115G3x, sc mTNF WT and sc mTNF Y86F3x were generated by gene synthesis (GeneArt). The individual chains are separated by a GGGGS (SEQ ID NO: 1) linker.

(22) scTNF-Nanobody Fusion Construction

(23) The coding sequence of the 1R59B Her2 nanobody was synthesized by PCR from the plasmid pHEN6-1R59B with the following primers:

(24) TABLE-US-00001 forward (SEQ ID NO: 2) 5′-GTCAAGATCTGGCGGTTCGGCGGCCGCAATGGCCCAGGTGCAG CTGCAG-3′, reverse (SEQ ID NO: 3) 5′-CAGTTCTAGATTACTTATCGTCGTCATCCTTGTAATCCGAACC GCCGTCCGGAGAGGAGACGGTGAC-3′.

(25) This PCR introduces a GGS in between a BglII and NotI site at the amino terminus and a FLAG tag at the carboxy terminus of the 1R59B nanobody. The PCR product was digested with BglII and XbaI. The pMK-RQ-sc hTNF WT, pMK-RQ-sc hTNF Y87Q3x and pMK-RQ-sc hTNF I97A3x were digested with NdeI and BglII. The digested PCR product and synthetic gene fragments were cloned into NdeI-XbaI digested pMET7 SIgK-HA-leptin vector to obtain pMET7 SIgK-HA-sc hTNF WT-6xGGS-1R59B-FLAG, pMET7 SIgK-HA-sc hTNF Y87Q3x-6xGGS-1R59B-FLAG and pMET7 SIgK-HA-sc hTNF 197A3x-6xGGS-1R59B-FLAG. The control vectors without the 1R59B nanobody were obtained by inserting the following annealed oligos containing the GGS and the FLAG tag in between BglII and XbaI instead of the PCR product:

(26) TABLE-US-00002 forward: (SEQ ID NO: 4) 5′ GATCTGGCGGTTCGGCGGCCGCAGATTACAAGGATGACGACGA TAAGTAAT 3′, reverse: (SEQ ID NO: 5) 5′CTAGATTACTTATCGTCGTCATCCTTGTAATCTGCGGCCGCCGA ACCGCCA3′.

(27) The control vector with only the 1R59B nanobody was obtained by inserting the following annealed oligos instead of the NdeI-sc hTNF-BglII fragment: forward: 5′-TATGATGTGCCCGACTACGCTGGCGGCAGCA-3′ (SEQ ID NO: 6), reverse 5′-GATCTGCTGCCGCCAGCGTAGTCGGGCACATCA-3′ (SEQ ID NO: 7). The length of the GGS linker was adjusted to a GGS linker of 13 repeats and 19 repeats by adding 7xGGS or 13xGGS repeats (made by gene synthesis, GeneArt) to the original 6xGGS in between the BglII and NotI site.

(28) A similar approach was used to obtain pMET7 SIgK-HA-sc hTNF WT-6x/13x/19xGGS-4.10-FLAG, pMET7 SIgK-HA-sc hTNF Y87Q3x-6xGGS-4.10-FLAG, pMET7 SIgK-HA-sc hTNF 197A3x-6x/13x/19xGGS-4.10-FLAG, pMET7 SIgK-HA-sc hTNF Y87Q1x 197A2x-6x/13x/19xGGS-4.10-FLAG, pMET7 SIgK-HA-sc hTNF Y87Q2x197A1x-6x/13x/19xGGS-4.10-FLAG, pMET7 SIgK-HA-sc hTNF WT-6x/13x/19xGGS-2HCD25-FLAG, pMET7 SIgK-HA-sc hTNF 197S3x-6x/13x/19xGGS-2HCD25-FLAG, pMET7 SIgK-HA-sc hTNF 197A3x-6x/13x/19xGGS-2HCD25-FLAG and pMET7 SIgK-HA-sc hTNF Y115A3x-6x/13x/19xGGS-2HCD25-FLAG.

(29) To obtain the individual trimerizing hTNF constructs, sc hTNF in pMet7-SIgK-HA-sc hTNF WT-GGS-4.10-Flag was replaced by NdeI-SalI digest of the PCR product obtained with the forward 5′-CATATGATGTGCCCGACTACGCTGGCGGCAGCAGCTCTAGAACCCCCAGCGATAAGCCTGTG-3′ (SEQ ID NO: 8) and the reverse primer 5′-GTCGACCAGGGCAATGATGCCGAAGT-3′ (SEQ ID NO: 9) on the plasmids pMet7-SIgK-His-hTNF WT or pMet7-SIgK-His-hTNF 197A. This resulted in the following vectors: pMet7-SIgK-HA-hTNF WT-6xGGS-4.10-Flag and pMet7-SIgK-HA-hTNIF197A-6xGGS-4.10-Flag.

(30) The nanobody-TNF fusion expression constructs with the NB N-terminally of individual trimerizing or single chain, human or mouse TNF were made in pMet7 and designed as such that each subunit is interchangeable through unique restriction sites: AgeI-nanobody-SalI-GGS linker-NotI-TNF-XhoI-His-XbaI.

(31) pGL3-(IL6-kB)3-fireflyluciferase was kindly provided by W. Vanden Berghe (Vanden Berghe et al., 1998).

(32) Production of the Nanobody-TNF Fusion Proteins for In Vitro Studies

(33) HekT cells were transfected with the protein fusion constructs using the standard calcium phosphate precipitation method. 48 hours after the transfection culture mediums were harvested and stored at −20° C. The concentration was determined with a commercial hTNF ELISA (DY210, R1D systems).

(34) Production of the Nanobody-scTNF Fusion Proteins for In Vivo Studies

(35) FreeStyle™ 293-F cells were transfected with the protein fusion constructs using the PEIpro™ transfection reagent (PolyPlus, Cat #115-375) according to the manufacturer's guidelines. The endotoxin content was in all preparations under the detection limit as assessed by a chromogenic Limulus Amebocyte Lysate Assay (Lonza, Cat #50-647U).

(36) Cell Lines

(37) Hek, HekT, Hek-mLR, MCF7, MCF7-hCD20, MCF7-mLR and B16Bl6-mCD20 cells were grown in DMEM supplemented with 10% FCS. The FreeStyle™ 293-F cell line was obtained from Invitrogen, Life Technologies (Cat #R790-07) and maintained in FreeStyle™ 293 Expression Medium from Gibco, Life Technologies (Cat #12338). The human breast cancer SK-BR-3 (ATCC: HTB-30) cell line was obtained from ATCC and maintained in McCoy's 5A medium supplemented with 10% FCS.

(38) The Hek-mLR cell line was generated as follows: Flp-In-293 cells (Invitrogen) were stably co-transfected with a plasmid containing the expression cassettes for mEcoR and neomycin resistance and with a pXP2d2-rPAP1-luci reporter construct (Eyckerman et al. 2001). Stable transfected clones were isolated in G418 (400 ug/ml)-containing medium and a clone was selected with high LIF (1 ng/ml)-induced luciferase activity. The expression vector pcDNA5/FRT containing the mLR was stably integrated in this cell line using the Flp-In recombinase reaction (Invitrogen) and after selection on hygromycin (100 μg/ml) for 10 days.

(39) The human breast cancer MCF7 (ATCC: HTB-22) cell line was obtained from ATCC. The MCF7-hCD20 and MCF7-mLR cell lines were generated as follows: MCF7 cells were stably co-transfected with a plasmid containing the expression cassette for hCD20 or mLR, and with a plasmid containing the neomycin resistance gene. Stable transfected cells were selected with G418 (1 mg/ml)-containing medium, followed by FACS sorting of hCD20- or mLR-expressing cells.

(40) The B16Bl6-mCD20 cell line was generated as follows: B16Bl6 cells were stably co-transfected with a plasmid containing the expression cassette for mCD20 and with a plasmid containing the neomycin resistance gene. Stable transfected cells were selected with G418 (2 mg/ml)-containing medium.

(41) The human breast cancer SK-BR-3 (ATCC: HTB-30) cell line was obtained from ATCC and maintained in McCoy's 5A medium supplemented with 10% FCS.

(42) Measurement of the Luciferase Activities

(43) TNF specific activities were measured by quantifying the luciferase activity under the control of the NF-KB promoter. Two days after transfection of the NF-KB luciferase reporter (pGL3-(IL6-KB)3-fireflyluciferase) by standard calcium phosphate precipitation method, cells were stimulated for 6 h with targeted or control sc hTNF. Lysates were prepared (lysis buffer: 25 mM Tris, pH 7.8, 2 mM EDTA, 2 mM dithiothreitol, 10% glycerol, 1% Triton X-100), and 35 μl of luciferase substrate buffer (20 mM Tricine, 1.07 mM (MgCO3)4Mg(OH)2.5H2O, 2.67 mM MgSO4.7H2O, 0.1 mM EDTA, 33.3 mM dithiothreitol, 270 μM coenzyme A, 470 μM luciferin, 530 μM ATP, final pH 7.8) was added per 50 μl of lysate. Light emission was measured for 5 s in a TopCount chemiluminescence counter (Packard).

(44) Quantitative RT-PCR

(45) The expression of the TNF inducible gene IL-6 was quantified by RT-PCR relatively to HPRT in SK-BR-3 cells treated for 6 hours with 500 ng/ml of targeted or control sc hTNF. Total RNA was purified with RNeasy columns (Qiagen) and equal amounts of RNA (0.5 μg) were used for reverse transcription using the Primescript RT Reagent kit (Takara Bio, Shiga, Japan), following the manufacturer's instructions. The 10-fold diluted cDNA was added to an RT-QPCR mixture containing 1×SYBR Green I master mix (04 887 352 001, Roche) and 1 nM gene-specific primers. Assays were performed in triplicate on a LightCycler 480 Real-Time PCR System thermocycler (Roche Applied Science), and the results were analyzed using the AACT method. The following primers were used:

(46) TABLE-US-00003 HPRT forward: (SEQ ID NO: 10) 5′TGACACTGGCAAAACAATGCA3′; HPRT reverse: (SEQ ID NO: 11) 5′GGTCCTTTTCACCAGCAAGCT3′; IL-6 forward: (SEQ ID NO: 12) 5′GACAGCCACTCACCTCTTCA3′; IL-6 reverse: (SEQ ID NO: 13) 5′AGTGCCTCTTTGCTGCTTTC3′.

(47) Toxicity Analysis on MCF7 Cells

(48) TNF-specific activities were also measured by assessing the cellular toxicity on MCF7 cells. 1000 cells were plated in a black 96-well plate and 24 hours later stimulated with the different TNF constructs. After 48-72 hours, the number of viable cells was determined using the CellTiter-Glo Luminescent Cell Viability Assay (Promega Cat #G7570) according to the manufacturer's guidelines.

(49) In Vivo Toxicity Analysis

(50) To assess hTNF toxicity in vivo, female 8 weeks old C57BL/6J mice (purchased from Charles River, France) were injected intraperitoneally with 500 ng rhTNF or sc hTNF-nanobody fusion proteins in combination with 10 mg D-Galactosamine (diluted in LPS-free PBS, injected in a volume of 500 μl). Morbidity was monitored by measurement of peripheral (rectal) body temperature. n=2-4 per fusion protein.

(51) To evaluate mTNF toxicity in vivo, mice were injected intravenously with 10, 35, 100 or 200 μg sc mTNF-nanobody fusion proteins (injected volume 200 μl, dilution in LPS-free PBS). Morbidity was monitored by measurement of peripheral (rectal) body temperature. n=2 per dose, per fusion protein, except for 200 μg (n=1).

(52) In Vivo Anti-Tumor Studies

(53) Female C57BL/6J mice of 8 weeks old were shaved and inoculated with 6×10.sup.5 B16Bl6-mCD20 tumor cells subcutaneously in the back (day 0). Treatment was started when the product of the largest perpendicular diameters was approximately 50 mm.sup.2 (on day 10). PBS or 35 μg nanobody-sc mTNF fusion proteins were administered for 8 consecutive days (day 10-17, indicated in FIG. 16A as a grey bar) via paralesional injection (subcutaneous injection near the tumor site but outside the tumor nodule). Tumors were measured daily with a caliper and are shown as mean±SEM. Morbidity was monitored by daily measurement of body weight and temperature. n=5 per treatment.

Example 1: The Sc hTNF-Nanobody Fusion Proteins

(54) FIG. 1 shows a schematic representation of the sc hTNF-nanobody fusion proteins either with the nanobody N- or C-terminally of sc hTNF.

Example 2: Targeting TNF Activity on mLR-Expressing Hek Cells

(55) The induction of NF-KB luciferase reporter activity upon TNF stimulation was tested in HekT cells and in Hek cells that express the murine leptin receptor (Hek-mLR). As shown in FIG. 2A, WT sc hTNF-induced NF-KB induction is completely (>1000-fold) or partly (100-fold) abrogated by the Y87Q3x or the I97A3x mutation, respectively. Moreover, in HekT cells that do not express the mLR, all sc hTNF constructs (WT, Y87Q3x and I97A3x) induce similar NF-KB activity independently of the fusion to the mLR nanobody (FIG. 2A). In contrast, coupling to the mLR nanobody is able to restore NF-KB induction of sc hTNF I97A3x in Hek cells that express the mLR to a similar extent as WT sc hTNF (FIG. 2B). We estimated that cells expressing the mLR are 100-fold more sensitive than parental HekT cells to the nanobody-coupled sc hTNF I97A3x. In contrast, the triple Y87Q mutation did not show any rescue effect of TNF responsiveness in Hek-mLR cells compared to HekT cells (FIG. 2B).

Example 3: Comparison of Different Mutant Combinations and Different Linker Lengths

(56) In order to optimize the constructs, sc hTNF constructs with different mutations in the individual chains were tested, as well as different linker lengths between the sc hTNF and the targeting moiety. The results are summarized in FIGS. 3, 4 and 5. sc hTNF I97A3x and sc hTNF Y87Q1x I97A2x do not show activity on Hek cells that do not express the leptin receptor, but have a clear dose dependent activity when targeted to the leptin receptor.

Example 4: Targeting TNF Activity on Her2-Expressing Hek Cells

(57) We generated fusions protein using the α-Her2 nanobody 1 R59B and sc hTNF WT, sc hTNF Y87Q3x or sc hTNF I97A3x. The linker between the nanobody and sc hTNF was either 6xGGS or 19xGGS. These molecules were tested on the Her2-overexpressing SK-BR-3 breast cancer cell line for the induction of the IL-6 TNF-inducible gene as determined relatively to HPRT by quantitative RT-PCR.

(58) FIG. 6 shows the fold induction of IL-6 mRNA upon sc hTNF treatment (500 ng/ml) compared to IL-6 mRNA levels in untreated cells and cells stimulated with the Her2 nanobody. In correspondence to the transcriptional activation of NF-KB, we observe that Y87Q3x mutation completely prevents TNF-induced IL-6 production while sc hTNF I97A3x can still induce IL-6 production but to a lesser extent than WT sc hTNF. When sc hTNF is fused to the nanobody less IL-6 mRNA is produced. This could be due to steric hindrance as the effect is more pronounced with the 6xGGS linker compared to the longer 19xGGS linker where there is likely more flexibility. By coupling sc hTNF I97A3x to the Her2 nanobody the induction of IL-6 can be restored to similar levels as WT sc hTNF coupled to the nanobody through the corresponding linker. In contrast, specific targeting of the more severe Y87Q3x sc hTNF mutant to Her2-expressing cells cannot restore the IL-6 inducing property of sc hTNF.

Example 5: Comparing the Toxicity of hTNF Mutants on MCF7 Cells

(59) Because of the relatively high residual activity of I97A3x mutant sc hTNF, we searched for further mutations by measuring the toxicity of different individual trimerizing hTNF mutants as luciferase activity in MCF7 cells. The activity of the mutants relative to WT individual trimerizing TNF is shown in FIG. 7. Most mutations do not affect the TNF activity drastically (>1% of WT) and might be less promising for the development of targeted constructs because of their possibly remaining substantial toxicity. We are more interested in mutations that (almost) completely abrogate TNF function. The use of null mutations (<0.1% of WT) results in targeted constructs that do not have side effects but that have as a possible drawback that reactivation upon targeting is less easily accomplished. The mutations that have some residual activity (0.02-5% of WT, particularly 0.1%-1% of WT) have a better chance of being reactivated whilst not being toxic. 6 different mutations covering an activity range between 0.02 and 5% of individual trimerizing WT TNF were selected for the development of the targeted modified cytokines: Y87Q (0.02%), 197S and Y115A (0.2%), Y87F (0.5-1%), Y115G (1-2%) and 197A (2-5%).

Example 6: Targeting TNF Activity on mLR-Expressing MCF7 Cells

(60) The toxicity of mLR NB-targeted TNF was assessed on MCF7 and MCF7-mLR cells. Different mutations (197A3x, I97S3x and Y115A3x) were tested as well as different linkers between sc hTNF and the mLR NB (6xGGS, 13xGGS, 19xGGS). As shown in FIG. 8A, toxicity is reduced 20-fold by the I97A3x mutation and 500-fold by the I97S3x and Y115A3x mutation, which is similar to what we observed for individual trimerizing TNF (FIG. 7). Moreover, in MCF7 cells that do not express the mLR, fusion to the mLR NB does not alter the activity of WT or mutant sc hTNF (FIG. 8A), while this fusion reactivates all sc hTNF mutants on MCF7-mLR cells (FIG. 8B). We estimated that cells expressing the mLR are 100-fold more sensitive than parental MCF7 cells to the NB-coupled I97S3x and Y115A3x sc hTNF. However, these targeted modified TNFs are still about 20-fold less active than WT sc hTNF. In contrast, the I97A3x targeted modified TNF is restored to WT activity levels on MCF7-mLR cells, which corresponds to a 20-fold reactivation (FIG. 8B).

Example 7: Targeting TNF Activity on hCD20-Expressing MCF7 Cells

(61) To assess the effect of other targeting moieties for the targeting of modified TNF, we replaced the mLR NB in the constructs of Example 6 with the hCD20 NB and tested their toxicity on MCF7 cells and MCF7 cells that express hCD20 (MCF7-hCD20). The results are shown in FIG. 9. As expected, mLR NB and hCD20 NB targeted modified TNFs behave similarly on parental MCF7 cells (FIGS. 8A & 9A).

Example 8: Targeting TNF Activity on hCD20-Expressing Cells with a Different hCD20 NB Fusion Set-Up

(62) We tried to improve the hCD20 NB-TNF constructs by placing the NB in front instead of after sc hTNF. We also tested 2 additional, less drastic mutations (Y87F3x and Y115G3x, FIG. 7). The MCF7 and MCF7-hCD20 toxicity studies with these constructs are shown in FIG. 10. Sc hTNF coupled to hCD20 NB exerts the same toxicity on MCF7 cells as the corresponding mutant coupled to the control BclI10 NB (FIG. 10A), and the level of activity is similar as to what we observed for the individual trimerizing TNF mutants (FIG. 7). This reduced toxicity of the mutants is (partially) reverted upon hCD20 targeting on the MCF7-hCD20 cells: hCD20 NB-coupled modified TNF give a 10-fold (Y115G3x), 15-fold (Y87F3x), 100-fold (19753x, Y115A3x) or even higher (Y87Q3x) increased activity compared to the corresponding BclI10 control NB-coupled sc hTNFs (FIG. 10B). In this experiment, when the hCD20 NB is placed at the carboxy-terminal end instead of the amino-terminal end of the sc hTNF the reactivation is less (FIG. 9B).

Example 9: Comparison of Different Mutant Combinations

(63) Despite the fact that the difference of targeted modified TNF versus non-targeted modified TNF is at least a 100-fold, some mutations show lower rescued activity than WT activity levels (Y87Q3x) which might affect its anti-tumor effects. Alternatively, some mutations still have some residual activity (19753x and Y115A3x) which might lead to some (systemic) toxicity when used in vivo. To overcome these potential drawbacks, we tested additional constructs by mutating different residues in the individual chains of sc hTNF in order to see whether the activity levels could thus be further modulated. As shown in FIG. 11, combining different mutations in the single chain can alter the residual activity on non-targeted cells and the level of reactivation upon targeting.

Example 10: Comparison of Targeted Individual Trimerizing TNF Versus Single Chain Modified hTNF

(64) To compare the efficiency of targeted individual trimerizing versus single chain TNF, WT or 197A hTNF was coupled C-terminally to the mLR NB as a monomer. Their toxicity was tested on MCF7 cells and on MCF7 cells that express the mLR (MCF7-mLR), and is shown in FIG. 12. Also in the individual trimerizing form, the 197A mutation is toxic on MCF7 cells but to a lesser extent than WT hTNF (FIG. 8A & FIG. 12A). Moreover, when coupled C-terminally to the mLR nanobody, individual trimerizing—but not single chain—TNF becomes less toxic on MCF7 cells, and this is the case both for WT and 197A (FIG. 8A & FIG. 12A). Most probably, the 3 nanobodies present in the hTNF trimer formed with the individual trimerizing TNF-mLR NB constructs are sterically hindering the binding of hTNF to its receptor. This reduced activity can, however, be reverted by targeting to the mLR on MCF7-mLR cells (FIG. 12B). Interestingly, this offers a further level of modulation of activity: one can combine different mutations, as well as use the sterical hindrance to influence residual activity and level of reactivation upon targeting.

(65) To address whether this is a general phenomenon, we coupled individual trimerizing WT and Y115A hTNF N-terminally to BclI10 or hCD20 nanobody and tested their toxicity on MCF7 and MCF7-hCD20 cells. As shown in FIG. 13A, this coupling does not affect the toxicity of individual trimerizing WT or Y115A3x hTNF. Moreover, upon targeting, individual trimerizing Y115A3x hTNF becomes as active as non-targeted individual trimerizing WT hTNF (FIG. 13B).

Example 11: Assessment of In Vivo Toxicity of Targeted Modified hTNF

(66) To evaluate the toxicity of hTNF mutants preclinically is not evident, since TNF displays a remarkable species specificity in mice. In contrast to mTNF, hTNF only induces lethality at extremely high doses (Brouckaert et al. 1992). Although the reason for this species specificity was long thought to be caused by hTNF not interacting with the murine TNF-R2, pharmacokinetic studies have shown that hTNF is cleared much faster than mTNF in mice and that the consequential limited hTNF exposure is responsible for its lack of morbidity (Ameloot et al. 2002).

(67) Nevertheless, when treated with a sensitizing agent such as D-galactosamine, species specificity is abolished and extremely low doses 500 ng) of hTNF are equally lethal as mTNF (Broeckaert et al., 1992). To assess the in vivo toxicity of the various targeted modified hTNFs, we therefore injected mice intraperitoneally with 500 ng of either recombinant (r) hTNF, sc hTNF WT or sc mutant hTNF (Y87Q3x or Y115A3x). The sc WT and modified hTNF were coupled N-terminally to either BclI10 or to hCD20 NB. As shown in FIG. 14, sc WT hTNF is at least as toxic as rhTNF, causing severe hypothermia and mortality within 10 h after injection. Targeted modified hTNF Y87Q3x and Y115A3x did not cause any signs of morbidity (pilo-erection, tremor, lethargy, loss of grooming or drop in body temperature; see FIG. 14A for the latter).

Example 12: Assessment of In Vivo Toxicity and Anti-Tumor Effect of Targeted Modified mTNF

(68) As already mentioned, in vivo toxicity of hTNF cannot be easily studied in mice. Therefore, as well as because of anticipated anti-tumor experiments in immunocompetent syngeneic mice, we decided to mutate residues of mTNF homologous to the ones we selected for hTNF (see example 5). As illustrated in FIG. 15, BclI10NB-sc mTNF WT caused severe morbidity (FIG. 15A) and 100% mortality (FIG. 15B) when injected intravenously in doses as low as 10 μg. In contrast, BclI10NB-sc mTNF Y86F3x did not induce mortality (FIG. 15B) nor cause any signs of toxicity (pilo-erection, tremor, lethargy, loss of grooming or drop in body temperature; see FIG. 15A for the latter), not even when injected as an intravenous bolus of 200 μg.

(69) Nevertheless, when injected daily paralesionally in a dose of 35 μg in B16Bl6-mCD20-tumor bearing mice, nanobody-coupled sc mTNF Y86F3x could still reduce/prevent tumor growth, especially when targeted to mCD20 (FIG. 16A). The effect of non-targeted mutant TNF on tumor growth (FIG. 16A) is due to the high dose (35 μg) used, as lower doses more closely mimic PBS-treated animals (data not shown). Daily treatment with the NB-sc mTNF Y86F3x did not cause any signs of morbidity or mortality, while tumor-bearing mice treated with NB-sc mTNF WT succumbed after 1 or 2 injections (FIG. 16B).

REFERENCES

(70) Ameloot, P., Takahashi N., Everaerdt, B., Hostens, J., Eugster, H. P., Fiers, W., and Brouckaert, P. (2002). Bioavailability of recombinant tumor necrosis factor determines its lethality in mice. Eur J Immunol, 32, 2759-65. Arap, W., Pasqualini, R. and Ruoslahti, E. 51998). Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science, 279, 377-380. Ashkenazi, A. & Herbst, R. S. To kill a tumor cell: the potential of proapoptotic receptor agonists. J Clin Invest 118, 1979-1990 (2008). Blake, A. W., McCartney, L., Flint, J., Bolam, D. N., Boraston, A. B., Gilbert, H. J. and Knox, J. P. (2006) Understanding the biological rationale for the diversity of cellulose-directed carbohydrate-binding molecules in prokaryotic enzymes. J. Biol. Chem. 281, 29321-29329. Boschert, V. et al. Single chain TNF derivatives with individually mutated receptor binding sites reveal differential stoichiometry of ligand receptor complex formation for TNFR1 and TNFR2. Cell Signal 22, 1088-1096 (2010). Brecht et al., Peptide immobilization and characterization of binding specificity. J Biol Chem 268: 15425-15434 Brouckaert, P., Libert, C., Everaerdt, B. and Fiers W (1992). Selective species specificity of tumor necrosis factor for toxicity in the mouse. Lymphokine Cytokine Res, 11, 193-6. Brown, K. C. (2010). Peptidic tumor targeting agents; the road from phagfe display selections to clinical applications. Curr. Pharm. Des. 16, 1040-1054. Daburon, S. et al. Functional characterization of a chimeric soluble Fas ligand polymer with in vivo anti-tumor activity. PLoS One 8, e54000 (2013). de Bruyn, M., Bremer, E. & Helfrich, W. Antibody-based fusion proteins to target death receptors in cancer. Cancer Lett 332, 175-183 (2013). Dimitrov, D. S. (2009) Engineered CH2 domains (nanoantibodies). mAbs 1, 26-28. Eyckerman, S., Waelput, W., Verhee, A., Broekaert, D., Vandekerckhove, J., and Tavernier, J. (1999). Eur. Cytok. Netw. 10, 549-559. Falschlehner, C., Ganten, T. M., Koschny, R., Schaefer, U. & Walczak, H. TRAIL and other TRAIL receptor agonists as novel cancer therapeutics. Adv Exp Med Biol 647, 195-206 (2009). Fox, N. L., Humphreys, R., Luster, T. A., Klein, J. & Gallant, G. Tumor Necrosis Factor-related apoptosis-inducing ligand (TRAIL) Receptor-1 and Receptor-2 agonists for cancer therapy. Expert Opin Biol Ther 10, 1-18 (2010). Galle, P. R. et al. Involvement of the CD95 (APO-1/Fas) receptor and ligand in liver damage. J Exp Med 182, 1223-1230 (1995). Gaur, U. & Aggarwal, B. B. Regulation of proliferation, survival and apoptosis by members of the TNF superfamily. Biochem Pharmacol 66, 1403-1408 (2003). Ghahroudi A. M., Desmyter, A., Wyns, L., Hamers, R., Muyldermans, S. (1997). Selection and identification of single domain antibody fragments from camel heavy-chain antibodies. FEBS Lett. 414, 521-6 Gregorc, V. et al. Phase Ib study of NGR-hTNF, a selective vascular targeting agent, administered at low doses in combination with doxorubicin to patients with advanced solid tumours. Br J Cancer 101, 219-224 (2009). Hehlgans, T. & Pfeffer, K. The intriguing biology of the tumour necrosis factor/tumour necrosis factor receptor superfamily: players, rules and the games. Immunology 115, 1-20 (2005). Huang, Y. & Sheikh, M. S. TRAIL death receptors and cancer therapeutics. Toxicol Appl Pharmacol 224, 284-289 (2007). Johnstone, R. W., Frew, A. J. & Smyth, M. J. The TRAIL apoptotic pathway in cancer onset, progression and therapy. Nat Rev Cancer 8, 782-798 (2008). Koivunen, E., Wang, B. and Ruoslahti, E. (1994). Isolation of a highly specific ligand for the α5β1 integrin from a phage library. J. Cell. Biol. 124, 373-380. Kolmar, H. (2008) Alternative binding proteins: biological activity and therapeutic potential of cysteine-knot miniproteins. FEBS J. 275, 2684-2690. Lejeune, F. J., Lienard, D., Matter, M. & Ruegg, C. Efficiency of recombinant human TNF in human cancer therapy. Cancer Immun 6, 6 (2006). Li, M. et al. Phase II multicenter, randomized, double-blind study of recombinant mutated human tumor necrosis factor-alpha in combination with chemotherapies in cancer patients. Cancer Sci 103, 288-295 (2012). Liu, Y. et al. The antimelanoma immunocytokine scFvMEL/TNF shows reduced toxicity and potent antitumor activity against human tumor xenografts. Neoplasia 8, 384-393 (2006). Loetscher, H., Stueber, D., Banner, D., Mackay, F. & Lesslauer, W. Human tumor necrosis factor alpha (TNF alpha) mutants with exclusive specificity for the 55-kDa or 75-kDa TNF receptors. J Biol Chem 268, 26350-26357 (1993). Nygren, P-A. (2008) Alternative binding proteins: affibody binding proteins developed from a small three-helix bundle scaffold. FEBS J. 275, 2668-2676. Ogasawara, J. et al. Lethal effect of the anti-Fas antibody in mice. Nature 364, 806-809 (1993). Pardon, E., Laeremans, T., Triest, S., Rasmussen, S. G. F., Wohlkonig, A., Ruf, A., Muyldermans, S., Hol, W. G. J, Kobilka, B. K. and Steyaert, J. (2014). A general protocol for the generation of Nanobodies for structural biology. Nature Protocols 9, 674-693 Roberts, N. J., Zhou, S., Diaz, L. A., Jr. & Holdhoff, M. Systemic use of tumor necrosis factor alpha as an anticancer agent. Oncotarget 2, 739-751 (2011). Roccaro A M, Hideshima T, Raje N, Kumar S, Ishitsuka K, Yasui H, Shiraishi N, Ribatti D, Nico B, Vacca A, Dammacco F, Richardson P G, Anderson K C. Bortezomib mediates antiangiogenesis in multiple myeloma via direct and indirect effects on endothelial cells. Cancer Res. 2006; 66(1):184-91. Scatchard G. Ann New York Acad Sci 1949; 51, 660-72. Schmidt T, Carmeliet P. Angiogenesis: a target in solid tumors, also in leukemia? Hematology Am Soc Hematol Educ Program. 2011; 2011:1-8. Siegemund, M. et al. Superior antitumoral activity of dimerized targeted single-chain TRAIL fusion proteins under retention of tumor selectivity. Cell Death Dis 3, e295 (2012). Skerra, A. (2008) Alternative binding proteins: anticalins—harnessing the structural plasticity of the lipocalin ligand pocket to engineer novel binding activities. FEBS J. 275, 2677-2683. Stump, M. T., Binz, H. K., Amstutz, P. (2008) DARPins: a new generation of protein therapeutics. Drug iscov. Today 13, 695-701. Tramontano, A., Bianchi, E., Venturini, S., Martin, F., Pessi, A and Sollazzo, M. (1994) The making of the minibody: an engineered beta-protein for the display of confromationally constrained peptides. J. Mol. Recognition 7, 9-24. Vanden Berghe, W. et al. p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways are required for nuclear factor-kappaB p65 transactivation mediated by tumor necrosis factor. J Biol Chem 273, 3285-3290 (1998). Vaneycken, I. et al. Preclinical screening of anti-HER2 nanobodies for molecular imaging of breast cancer. FASEB J 25, 2433-2446 (2011). van Horssen, R., Ten Hagen, T. L. & Eggermont, A. M. TNF-alpha in cancer treatment: molecular insights, antitumor effects, and clinical utility. Oncologist 11, 397-408 (2006). Wang, H., Yan, Z., Shi, J., Han, W. & Zhang, Y. Expression, purification, and characterization of a neovasculature targeted rmhTNF-alpha in Escherichia coli. Protein Expr Purif 45, 60-65 (2006). Welti J, Loges S, Dimmeler S, Carmeliet P. Recent molecular discoveries in angiogenesis and antiangiogenic therapies in cancer. J Clin Invest. 2013; 123(8):3190-200. Yang, Y. H., Rajaiah, R., Ruoslahti, E. and Moudgil, K. D. (2011). Peptides targeting inflamed synovial vasulature attenuate autoimmune arthritis. PBNAS 108, 12857-12862. Zabeau, L. et al. Selection of non-competitive leptin antagonists using a random nanobody-based approach. Biochem J 441, 425-434 (2012).