Synthesis and evaluation of peptide and nucleic acid based Toll-like receptor ligands
Weterings, J.J.
Citation
Weterings, J. J. (2008, November 27). Synthesis and evaluation of peptide and nucleic acid based Toll-like receptor ligands. Bio-organic Synthesis, Leiden Institute of Chemistry, Faculty of Science, Leiden University. Retrieved from https://hdl.handle.net/1887/13284
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/13284
Note: To cite this publication please use the final published version (if applicable).
Chapter 6
Summary and Future Prospects
This thesis describes the design, synthesis and immunological evaluation of ligands and conjugates aimed at activation of Toll-like receptor 2, 7 and 9 (TLRs 2, 7 and 9). The role of the TLRs in human immunity is discussed in the introductory Chapter 1, with the focus on some well-defined natural and synthetic ligands currently studied and that act on one of the eleven TLRs known to exist in man. Chapter 1 describes a selection of natural and synthetic TLR ligands available to date with a focus on synthetic, well-defined compounds. In Chapter 2 the design and preparation of a series of novel 7-hydro-8-oxo-adenines and their TLR agonistic properties are described. Literature compound 1, a potent TLR7 agonist, formed the starting point for these studies and it was found that introducing an azide-containing side chain to the purine core (as in 2, being a representative example, Figure 6.1) gave derivatives, that, although less active than the parent compounds, still possessed considerable TLR7 agonizing properties. The azide functionality in derivative 2 and its homologues open the way to a variety of new potential TLR7 ligands.
NH2
O NH HN
O AA
23
H2N
O N
N N
O N N
NH N
NH2
O
O O
AA
n = DEVSGLEQLESIINFEKLAAAAK 3
O N N
NH N
NH2
O
O O
N3
O N N
NH N
NH2
O O
1 2
Figure 6.1
Summary, General Evaluation and Future Prospects
Although presentation of the antigenic peptide, SIINFEKL, proceeded more efficiently using these constructs compared to the separate TLR-L and peptide parts, TLR7 dependent cytokine production was abolished. Several factors may be at the basis of this outcome. Literature describes that the optimal length of the C2-spacer in the TLR7-L is four to six atoms. If the peptide is considered as part of this spacer after the Huisgen cycloaddition, this optimal length is surpassed to a huge extent. Second, aggregation of the peptide conjugate can bury the TLR7 ligand inside the construct, so that it is no longer able to bind to TLR7 after uptake. Third, the uptake of the construct itself might be hindered, thereby creating a concentration of TLR7-L too low to effectively trigger its receptor. Fourth, the position of ligation between the TLR7-L and the antigenic peptide remains a point of consideration. Last, also the concept of the covalent bond may be reconsidered, and constructs able to enter the target cell as a conjugate and then dissociate under physiological conditions into the separate TLR-L and antigenic peptide should not be discarded. Future research making use of more advanced constructs may give a more detailed insight into these issues, and ultimately should give an answer to the question whether it is at all feasible to use modified 7-hydro-8-oxo-adenines as TLR7 ligands in synthetic vaccine constructs. Three such advanced constructs are depicted in Figure 6.2.
As a first example construct 6 is devised to connect the TLR7 pharmacophore to the antigenic peptide via meta-substitution of the aryl functionality at N9. Retrosynthetically, construct 6 is available from peptidic acetylene 5 and iodophenyl derivative 4 by means of Sonogashira cross-coupling. The synthesis of 4 can be accomplished following the route outlined in the experimental part at the end of this chapter, and preliminary studies indicate that the execution of the key Sonogashira reaction is feasible on solid support. Future research is required to establish the validity of this approach, and whether construct 6 induces TLR7-dependent cytokine production and leads to the required antigen presentation.
OH O AA
18 HN
O NH S O
N N
HN H2N N
O
S
S KKKK H N O N H S O
C15H31 O O C15H31
O
O C15H31
OH O AA
18 HN
NH O N
N H N
N
H2N O
SKKK K H N O N H S O
C15H31 O C15H31 O
O
O C15H31
NH O
N N N
8 7
A A 18 HN
O
OH O AA
18 HN
O
O
N N
HN H2N N
O
5
6 AA
n
x n = 18 = DEVSGLEQLESIINFEK L
O
+ N
N N
H N NH2
O
O
O 4
I
O O
O
Figure 6.2
Two alternative constructs are conjugates 7 and 8. Both feature a Pam3Cys moiety as a TLR2 ligand. In compound 7 the TLR2 ligand is conjugated through a 1,4-triazole linkage, whereas the connection of the 7-hydro-8-oxo-adenine derivative in 8 is made through a disulfide bond.
Thus, construct 7 is deemed stabile under the physiological conditions present in the endosomal and lysosomal compartments, whereas the disulfide bond in 8 would be reduced in these compartments. These constructs will give insight in whether A) poor uptake of 7-hydro- 8-oxo-adenine constructs such as compound 3 prohibit cytokine production (then compound 7 should give a positive result) or B) whether the TLR7 ligand needs to be liberated from the construct prior to binding and agonizing TLR7 (here compound 8, but not 7, gives TLR7- dependent cytokine production).
Summary, General Evaluation and Future Prospects
CCATGACGTTCCTGACGT 5'
O 3'
P O S
O-
O N NH
O
O O
P O
S O- O
HO N
NH O
O
O P S -O
O 9
H
N DEVSG LEQLES IINFE S
N O
O
O HN
NH
NH O
OH O O
O
NH2+ SO3- S O3- H2N
COO H
O
O C15H31 O O C15H31
O S
N H
O
*
C15H31 O
H N S KKKK
HN
O O H
N DEVSG LEQLES IINFEK L-OH S
N
O
O NH
O N N F B
F -
+
10
O C15H31 O O C15H31
O S
N H
O 2
C15H31 O
H N R'
11 1
O C15H31 O O C15H31
O S
N H
O 2
C15H31 O
H N R' 1
R S
12
5 6
Figure 6.3
Chapters 4 and 5 discuss the synthesis and evaluation of several peptide-TLR-L conjugates based on the TLR9 ligand CpG (see for a representative example fluorescent derivative 9) and TLR2 ligand Pam3Cys (representative example, structure 10, Figure 6.3). With these, the trafficking, localization and the effectiveness of antigen presentation in conjunction with cytokine production could be studied in some detail. One specific issue, addressed in Chapter 5, pertains the influence of the absolute stereochemistry of the chiral glycerol carbon in Pam3Cys constructs on TLR2 binding activity. With the aid of compounds 11 and 12 (with R’
encompassing the antigenic peptide) it was found that the unnatural S-configuration construct results in less effective cytokine production but similar antigen presentation.
The results described in Chapter 4 and 5 demonstrate the complexity of the uptake / the TLR binding / the processing and antigen presentation, events that need to take place for both antigen presentation and cytokine production being the minimal requirements of any synthetic construct for potential vaccine development. Although some questions are answered by
making use of constructs such as described in this thesis, many more remain unanswered and may be tackled by yet more advanced constructs. An important issue in immunology is whether TLR-ligands bind to their TLR partner directly or whether mediators and/or cofactors are involved. Examples of such constructs are the conjugates featuring a TLR2 (13) or a TLR9 ligand (14) provided with both a biotin and a photocrosslinker. Photocrosslinking experiments using 13 and 14 (Figure 6.4) should give insight into this.
N H NH2
O H N O
HN
O HO N H S O O
O
C15H31 C15H31
O
C15H31 O
NH O
O HN
O 4
NH2
HN NH
HN
NH2 O
O NH
O
O
NH N O 7
N O
O S P S
O- CpG DNA
3' 5'
13
14 S
HN NH O
H H
O H N
O
S HN NH
O
H H
O HN
O
Figure 6.4
Altogether, the tailor-made constructs, aimed at TLR2, 7 and 9 described in this thesis present a first step towards the design of synthetic, well-defined vaccines that recruit these receptors in generating a directed immune response. It is obvious that much research is needed to make such a putative synthetic vaccine come true. However, the discovery of the TLR family and their role in immunity, the potential to formulate specific and well-derived bioconjugates that act on the TLRs, and the continuous development of synthetic methodologies that enables the preparation of such defined bioconjugates support the idea that such synthetic vaccines are indeed feasible. Future research will show whether TLR2, 7 or 9 are the right TLRs to target, or whether it is better to target other members of this family, or even combinations of TLRs.
The results presented in this thesis and stemming from several literature reports demonstrate that, if selected with care, covalent linking a TLR-L with a peptide epitope gives enhancement of both cytokine production and antigenic peptide presentation, the two hallmarks of synthetic vaccine development of this type. It is not unlikely that future synthetic vaccines, different in
Summary, General Evaluation and Future Prospects
Experimental Section
9-meta-iodobenzyl-2-chloro-adenine
2-chloroadenine (500 mg, 2.96 mmol) was dissolved in a 1M solution of TBAF in THF (5.92 mmol, 2eq.). m-iodo-benzylbromide (2eq., 5.92 mmol, 1.75 g.) was added and the mixture was allowed to stir at rt. When TLC indicated a complete reaction the solvent was removed in vacuo and the crude was subjected to column chromatography. A gradient of 10% EtOAc in Toluene to 100% EtOAc yielded the product as a white solid. Yield: 582 mg, 1.51 mmol (51%). 1H NMR (400 MHz, DMSO-d6): 8.25 (s, 1H, H8) 7.91 (bs, 2H, NH2), 7.24 (s, 1H, CH Arom.), 7.66 (d, 1H, CH Arom.), 7.25 (d, 1H, CH Arom.), 7.15 (t, 1H, CH Arom.), 5.30 (s, 2H, CH2Bn); 13C NMR (100 MHz, DMSO-d6): 156.8, 153.1, 150.5 (Cq), 149.4 (C-8), 139.2 (Cq), 136.5, 136.0, 130.9, 126.8 (C Arom.), 117.7, 95.0 (Cq), 45.4 (CH2Bn); IR: 3123, 1678, 1607, 1576, 1308, 1250. ESI-MS: m/z 386.0 [M+H]+. HRMS: C12H9N5ClI + H+ calculated. 385.96639, found 385.96635 (35Cl)
9-meta-iodobenzyl-2-methoxyethoxy-adenine
9-meta-iodobenzyl-2-chloro-adenine (200 mg, 0.52 mmol) was dissolved in fresh distilled ethyleneglycolmonomethylether (10 mL). NaH (5 eq., 2.61 mmol, 104 mg) was added gradually over 5 min. The reaction was heated to 850C and stirred overnight. The mixture was concentrated and the crude was subjected to column chromatography. A gradient of 1 to 4%
MeOH in DCM yielded a white powder. Yield: 157 mg, 0.37 mmol (71%). 1H NMR (200 MHz, CDCl3): 8.05 (s, 1H, H8), 7,75 (s, 1H, CH Arom.), 7.65 (d, 1H, CH Arom.), 7.32 (d, 1H, CH Arom.), 7.16 (bs, 2H, NH2), 7.14 (t, 1H, CH Arom.), 5.21 (s, 2H, CH2Bn), 4.32 (t, 2H, CH2), 3.61 (t, 2H, CH2), 3.27 (s, 3H, CH3); 13C NMR (100 MHz, CDCl3): 161.3, 156.8, 151.0, 139.6 (Cq), 139.4 (C-8), 136.4, 130.8, 127.2 (CH Arom.), 115.1, 94.9 (Cq), 70.3, 65.3 (CH2), 58.1 (CH3), 45.3 (CH2Bn); IR: 3323, 1626, 1589, 1387, 1323, 1066. ESI-MS: m/z 426.1 [M+H]+. HRMS: C15H16N5O2I + H+ calculated. 426.04214, found 426.04234.
9-meta-iodobenzyl-2-methoxyethoxy-8-bromo-adenine
9-meta-iodobenzyl-2-methoxyethoxy-adenine (200 mg, 0.47 mmol) was dissolved in DCM.
Bromine (500 μL) was added, the reaction was shielded from light and stirred for 1 hour. TLC indicated a quantitative conversion into the bromide. Excess bromine was removed by constant airflow and coevaporating the crude with CHCl3 (thrice). After loading of the impure product onto celite, column chromatography (15% Toluene/EtOAc to 5% MeOH/EtOAc) yielded a pure white powder. Yield: 141 mg, 0.28 mmol (60%). 1H NMR (400 MHz, DMSO-
d6): 7.69-7.65 (m, 2H, CH Arom), 7.49 (bs, 2H, NH2), 7.20-7.13 (m, 2H, CH Arom.), 5.21 (s, 2H, CH2Bn), 4.32 (t, 2H, CH2), 3.62 (t, 2H, CH2), 3.27 (s, 3H, CH3); 13C NMR (100 MHz, DMSO-d6): 161.3, 155.7, 152.3, 138.5 (Cq), 136.6, 136.0, 131.0, 126.6 (C Arom.), 123.7, 115.4, 95.0 (Cq), 70.2, 65.5 (CH2), 58.1 (CH3), 45.6 (CH2Bn); IR: 3420, 3315, 3186, 2924, 1634, 1591, 1393, 1313, 1173, 1097. ESI-MS: m/z 503.9 [M+H]+. HRMS: C15H15N5O2BrI + H+ calculated. 503.95266 (79Br), found 503.95245.
9-meta-iodobenzyl-2-methoxyethoxy-8-methoxy-adenine
The bromide (71.0 mg, 0.14 mmol) was dissolved in MeOH (10 mL), after which NaOMe (20 eq., 2.82 mmol, 152 mg) was added. The reaction was stirred at 650C overnight. After in vacuo removal of the solvent the crude was loaded onto celite. Column chromatography using a gradient of 1 to 5% MeOH in DCM yielded a white product.Yield: 62 mg, 0.14 mmol (97%). 1H NMR (400 MHz, DMSO-d6): 7.65 (s, 1H, CH Arom.), 7.63 (d, 1H, CH Arom.), 7.21 (d, 1H, CH Arom.), 7.14 (t, 1H, CH Arom.), 6.88 (bs, 2H, NH2), 5.00 (s, 2H, CH2Bn), 4.28 (t, 2H, CH2), 4.04 (s, 3H, 8-OCH3), 3.60 (t, 2H, CH2), 3.17 (s, 3H, CH3); 13C NMR (100 MHz, DMSO-d6): 159.9, 145.5, 153.3, 150.8, 139.2 (Cq), 136.3, 135.9, 130.9, 126.6 (C Arom.), 110.0, 94.9 (Cq), 70.3 (CH2), 65.2 (CH2), 58.1 (CH3), 56.9 (CH3), 43.0 (CH2Bn); IR:
3441, 3321, 2810, 1626, 1599, 1564, 1387, 1333, 1317, 1065, 1026. ESI-MS: m/z 456.0 [M+H]+. HRMS: C16H18N5O3I + H+ calculated 456.05271, found 456.05227.
9-meta-iodobenzyl-2-methoxyethoxy-7-hydro-8-oxo-adenine (4)
The methoxy compound (62 mg, 0.14 mmol) was dissolved in 5 mL concentrated HCl and allowed to stir overnight. After the solution was evaporated to dryness, 10 mL of milipore water was added. The whole was basified using sat NH4OH. The water layer was extracted with CHCl3 (thrice, 10 mL), dried (MgSO4), filtered and concentrated. Yield: 57 mg, 0.13 mmol (92%). 1H NMR (400 MHz, DMSO-d6): 10.0 (s, 1H, OH), 7.69 (s, 1H, CH Arom.), 7.66 (d, 1H, CH Arom.), 7.28 (d, 1H, CH Arom.), 7.14 (t, 1H, CH Arom.), 6.50 (bs, 2H, NH2), 4.81 (s, 2H, CH2Bn), 4.26 (t, 2H, CH2), 3.59 (t, 2H, CH2), 3.26 (s, 3H, OCH3); 13C NMR (100 MHz, DMSO-d6): 159.8, 152.1, 149.0, 147.8 139,7 (Cq), 136.2, 130.8, 127.0 (CH Arom.), 98.3, 94.8 (Cq), 70.2, 65.3 (CH2), 58.1 (CH3), 41.2 (CH2Bn). ESI-MS: m/z 442.0 [M+H]+. HRMS: C15H16N5O3I + H+ calculated 442.03706, found 442.03677.
