This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

Dendritic structures, being highly homogeneous and symmetric, represent ideal scaffolds for the multimerization of bioactive molecules and thus enable the synthesis of compounds of high valency which are e.g., applicable in radiolabeled form as multivalent radiotracers for ^{68}Ga. Thus, this solid phase-assisted dendron synthesis approach enables the fast and straightforward assembly of bioactive multivalent constructs for example applicable as radiotracers for

Dendritic structures, exhibiting highly branched and ideally well-defined and homogeneous structures, have extensively been studied as carriers for a large variety of bioactive compounds over several decades [

The standard procedure for the preparation of dendrimers is still the solution-phase synthesis as by this technique, very large dendritic structures can be synthesized on a large scale. However, this synthesis approach is often time-consuming due to the long reaction times required for a complete conversion of all functional groups and the cumbersome purification procedures that have to be implemented after each reaction step.

An alternative is the synthesis of dendrimers using a solid support, which exhibits the advantages of comparatively fast reaction kinetics (for each synthesis step, reaction times of several hours per reaction step are required, whereas in case of solution phase dendrimers, reaction times of days to weeks have to be applied to obtain substances of high homogeneity) and very efficient purifications which are executed by simple washing of the resin.

So far, several examples for the solid phase-assisted synthesis of dendrimers are available [

Such a modular dendrimer design can be of special interest if the dendrimers are to be used as drug carriers with a certain, optimal distance required between two drug moieties within the molecule. So was for example shown very recently that the distance between peptide moieties within a dendritic peptide multimer has a crucial influence on the achievable binding avidities of the multivalent peptides to their respective receptor [

Only very few examples of solid phase-supported syntheses of structurally flexible polylysine dendrimers have been described recently and in these systems, the structural flexibility was achieved by the introduction of different linkers [

Such dendritic scaffolds, offering optimal properties such as a symmetrically branched and highly homogeneous structure while at the same time enabling a tailored molecular design with variable distances between surface functionalities, are an ideally suited platform for example for the synthesis of peptidic multimers. Such multivalent bioactive compounds have shown to be highly favorable radiotracers for the diagnostic imaging of different cancer types when radiolabeled and applied in

To improve the existing synthesis pathways towards dendritic structures which are highly flexible in molecular design, are accessible by solid phase-supported synthesis protocols and can be furthermore applied in multimerization reactions and molecular imaging settings, we intended to develop a synthesis route which should fulfill the following requirements: (^{68}Ga.

For the synthesis of dendritic structures, two different processes can be applied: a divergent or a convergent synthesis approach (

Hence, a divergent route was followed for this solid phase-assisted dendron synthesis approach as this allows for the use of different building blocks in every synthesis step, thus a maximal modular structure of the resulting dendritic scaffolds as well as tailored distances between two dendron surface functionalities.

Schematic depiction of a divergent (

The dendron assembly on solid support thus comprises the following steps: (

Besides enabling a highly flexible synthesis and thus molecular structure, the principle design of the dendrons should follow the composition of conventionally synthesized PAMAM dendrons successfully used before for the synthesis of multivalent peptides [

Schematic depiction of the general structure of the dendritic scaffolds synthesized on solid support. The dendrons comprise terminal maleimide functionalities for the efficient Michael addition of arbitrary thiol-bearing molecules, symmetrical branching units, oligoethyleneglycols of differing lengths for modular molecular design and tailored distances between two adjacent dendron surface functionalities as well as the possibility to introduce a chelator or thiol functionality for radiometal labeling or biomolecule conjugation, respectively.

To obtain dendritic structures applicable as scaffolds for the multimerization of arbitrary thiol-bearing bioactive molecules while also enabling the introduction of a chelator for radiometal labeling, the first synthesis attempts comprised the conjugation of a lysine to the solid support and the subsequent conjugation of the branching units. For the first synthesis experiments, a standard rink amide resin with a loading of 0.74–0.79 mmol/g was used as solid support and the conjugation reactions were performed using standard Fmoc solid phase peptide synthesis protocols and reactant excesses of four equivalents of amino acid per amino functionality to be derivatized. As branching unit,

Schematic depiction of the reaction pathway for the synthesis of non-OEG-comprising dendritic structures on the example of a maleimide tetramer.

The resulting dimers and tetramers synthesized by this procedure could however be obtained in only low purities of the cleaved raw materials. These findings can be attributed to the steric hindrance exerted by the relatively bulky structures and their small distances to one another, impeding efficient conjugation reactions.

In the following was thus used the same conjugation chemistry but a low-loading rink amide resin with a loading of only 0.21–0.23 mmol/g. Furthermore, an OEG linker was implemented between the initially conjugated lysine or cysteine amino acid and the focal point of the dendrons. These measures resulted in significantly improved raw product purities of 83%

High purities of the cleaved raw materials are a prerequisite for a straightforward synthesis and also for reasonable isolated product yields and were thus in the following used as a measure for the synthesis efficiency.

In the following, the length of the OEG linker inserted between the initial amino acid and the focal point of the dendrons (_{1}, PEG_{3}, PEG_{5}, PEG_{7} and PEG_{11}; OEG linkers are denoted as “PEGs” due to their trade names). For the tetramers, the length of the applied OEG did not seem to have a crucial influence on the achievable product purities after cleavage as long as it exceeded the length of a PEG_{1} (

In contrast to these findings for tetramers, the highest maleimide octamer raw product purities obtained after cleavage were found using a PEG_{11} as initial

Analytical HPLC traces of the raw materials of a tetravalent maleimide synthesized on a standard rink amide resin comprising no OEG linker (_{5}-linker (

Observed product purities of the cleaved raw materials (

To evaluate if the reaction efficiency can be further optimized introducing additional OEG linkers in other positions of the dendrons (

In contrast to these findings for tetramers—and as shown before for the initially conjugated _{1}-linkers gave the best results when using the same linker length in every position (

Obtained product purities of the cleaved raw materials (

Analytical HPLC traces of the raw materials of octavalent maleimides synthesized comprising only one initial PEG_{1} (_{1}-linkers in each possible position of the dendron (

This negative effect of using long linkers in each position of the dendron scaffold may be attributable to the strongly increased flexibility when using long OEGs which on the one hand decreases the steric hindrance for the following reactions but on the other hand obviously also increases entropic or coiling effects hampering an efficient conjugation of following building blocks thus outbalancing the positive effects of linker introduction.

Overall, the highest reaction efficiencies were observed for octavalent maleimides containing a PEG_{11} linker between the initial amino acid and the focal point of the dendron and additional PEG_{1} linkers after the branching units. Interestingly, the introduction of an OEG linker before conjugating the surface maleimide functionalities did not give improved, but even deteriorated results (comparing results obtained for

Taken together, these results indicate that not only the steric hindrance between dendritic scaffold and resin can hamper the efficient conjugation reactions of building blocks but also the steric hindrance which is exerted by the dendritic structure itself if exceeding a certain size.

Taking these findings into account, maleimide hexadecimers were synthesized in the following and could be obtained in purities of the raw materials of up to 38% (

Obtained product purities of the cleaved raw materials (a); and schematic depiction of the structures of hexadecimeric maleimides

Besides the molecular design which was shown to exert a significant influence on reaction efficiencies and thus product yields, also other parameters turned out to be important to obtain optimal synthesis results. It could be observed that—with an increasing number of branching units within the dendrons and thus an increasing steric demand of the constructs—the conjugation times for the following building blocks had to be prolonged and furthermore depended on the structure of the synthon to be coupled. So could be shown that for linear building blocks such as OEG linkers and maleimido hexanoic acid, the optimal results could be obtained extending the reaction times from 30 min over 45 min and 60 min to 120 min, depending on the number of branching units and the position within the dendritic structure. In case of the branching amino acid

Besides reaction times, the reactant excess used also has a considerable influence on the achievable reaction efficiencies. Astonishingly, the best results were not obtained for a reactant excess of four equivalents of building block per reactive functionality as it is generally applied in Fmoc solid phase peptide synthesis. In contrast, two equivalents turned out to be optimal. Thus, for example it could be shown that during the synthesis of

All in all, it could be demonstrated that the synthesis of even comparably large multivalent structures of up to hexadecimers could be accomplished on solid support. Varying the structural design of the dendrons, the products could be obtained in reasonable yields and although showing preparation optima for certain structures, also deviant structures could be successfully synthesized. Furthermore, the preparation times for the solid phase-dendrons were between 1 and 3 days—depending on the generation of the dendritic structure—which is considerably faster than the equivalent dendron synthesis in solution. So could e.g., octameric structures be completely assembled on solid support within less than 24 h using this procedure whereas the corresponding PAMAM dendrons have to be synthesized over a reaction time of 30 days (over all reaction steps), not including the time required for the mandatory purifications, to obtain highly homogeneous products [

To show the applicability of the synthesized dendritic multivalent maleimide scaffold structures in the multimerization of thiol-bearing synthons and subsequent radiolabeling, different scaffolds were synthesized according to the afore-mentioned optimized reaction conditions. In principle, different multivalent structures could be of interest for an application in

Schematic depiction of the synthesized mono/multivalent maleimides

Concerning the introduction of DOTA into the dendritic scaffolds, it was observed that only slightly higher product purities of the raw materials (~4%) could be obtained when conjugating tris-

The maleimide-comprising scaffolds to be applied in the following multimerization reactions were obtained in comparably low isolated yields between 2.9% and 17.0% after HPLC purification although the product purities of the raw materials after cleavage were adequate (between 53% and 78%). Furthermore, the observed isolated yields did not depend on dendron size or complexity (e.g., higher yields for di- and tetramers and lower yields for octamers would have indicated a loss of material during HPLC purification increasing with the size of the constructs due to growing interactions with the column material). Thus, a possible explanation for this effect of relatively low isolated yields could be an incomplete cleavage of the products from the solid support although even prolonged cleavage times did not give improved results. Another possibility might be that not all functional amides on the resin are sufficiently accessible to participate in chemical reactions as it was observed that even the mass of the isolated raw materials considerably deviated from the substance amounts that were estimated from the specified theoretical loading of the resin. To further substantiate this latter assumption, the resin was weighed before and after the assembly of the multivalent systems. By these experiments, it was e.g., found during the synthesis of

The tetravalent and octavalent maleimides _{8}-DOTA (_{8}-DOTA (_{8}-DOTA (_{8}-SS-_{4}-DOTA (_{4}-DOTA (_{4}-SS-

Schematic depiction of the reaction of the multivalent maleimides

Another observation was that the conjugation of the thiol-bearing synthons to the multivalent maleimides could be performed in solution after cleavage of the dendritic systems from the solid support and purification of the scaffolds, but also on solid support, thus further reducing the synthesis efforts for the multivalent constructs. This was exemplarily studied on the tetravalent compounds

These results show the applicability of the multivalent maleimides synthesized on solid support for the multimerization of even structurally different building blocks.

As the aim of this study was to synthesize dendritic scaffolds that are not only applicable for the multimerization of varying thiol-bearing bioactive molecules but also for radiolabeling to be useful as radiotracers for ^{68}Ga using the fractioned elution methodology of ^{68}Ge/^{68}Ga generators [^{68}Ga-labeling conditions [^{64}Cu could be more appropriate but depends on the actual

The ^{68}Ga-radiolabeling of the model octamers ^{68}Ga-labeling of ^{68}Ga^{3+}.

All commercially available chemicals were of analytical grade and used without further purification. Resins for solid-phase synthesis, coupling reagents, Fmoc-protected amino acids, Fmoc-NH-PEG_{1}-COOH and Fmoc-NH-PEG_{3}-COOH were purchased from NovaBiochem (Schwalbach, Germany), 1-thio-β-D-galactose sodium salt, Fmoc-NH-PEG_{5}-COOH, Fmoc-NH-PEG_{7}-COOH and Fmoc-NH-PEG_{11}-COOH were obtained from Iris Biotech (Marktredwitz, Germany), 6-maleimidohexanoic acid and L-glutathione (reduced) were purchased from Sigma Aldrich (Schnelldorf, Germany),

The dendron scaffolds were synthesized on solid support by standard Fmoc solid-phase peptide synthesis using an amino acid excess of 2 eq. per functionality and a standard conjugation time of 30 min, a commercially available standard Rink Amide resin (for compounds _{α}-Fmoc-amino acids, Fmoc-_{ω}-PEG-amino acids and _{2}O (95:2.5:2.5) for 60 min if the scaffolds did not contain a protected DOTA building block or 2 h if they contained tris-

_{2}N-CO-Lys-PEG_{1}-Mal_{4}_{t} = 3.4 min; ESI-MS (^{3+} (calculated): 526.31 (526.31); [M+2H]^{2+} (calculated): 788.96 (788.96); [M+2Na]^{2+} (calculated): 810.94 (810.94).

_{2}N-CO-Lys-PEG_{3}-Mal_{4}_{t} = 3.5 min; ESI-MS (^{3+} (calculated): 560.33 (560.33); [M+2H]^{2+} (calculated): 839.99 (839.99); [M+H+Na]^{2+} (calculated): 850.98 (850.98); [M+2Na]^{2+} (calculated): 861.97 (861.97); [M+Na+K]^{2+} (calculated): 869.96 (870.03).

_{2}N-CO-Lys-PEG_{5}-Mal_{4}_{t} = 3.7 min; MALDI-MS (^{+} (calculated): 1766.55 (1767.03); [M+Na]^{+} (calculated): 1788.98 (1789.01); [M+K]^{+} (calculated): 1805.58 (1805.12).

_{2}N-CO-Lys-PEG_{7}-Mal_{4}_{t} = 3.8 min; ESI-MS (^{3+} (calculated): 619.04 (619.03); [M+2H]^{2+} (calculated): 928.05 (928.04); [M+H+Na]^{2+} (calculated): 939.04 (939.04); [M+2Na]^{2+} (calculated): 950.03 (950.03); [M+Na+K]^{2+} (calculated): 958.02 (958.08).

_{2}N-CO-Lys-PEG_{11}-Mal_{4}_{t} = 4.1 min; MALDI-MS (^{+} (calculated): 2031.36 (2031.19); [M+Na]^{+} (calculated): 2053.29 (2053.17).

_{2}N-CO-Lys-(PEG_{1})_{3}-Mal_{4}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, 6-maleimido-hexanoic acid: 1h; gradient: 20%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.9 min; ESI-MS (^{3+} (calculated): 816.46 (816.46); [M+2H+Na]^{3+} (calculated): 823.78 (823.78).

_{2}N-CO-Lys-(PEG_{3})_{3}-Mal_{4}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, 6-maleimido-hexanoic acid: 1 h; gradient: 20%–55% MeOH + 0.1% FA in 5 min, R_{t} = 3.8 min; ESI-MS (^{4+} (calculated): 791.21 (791.21); [M+Na+3H]^{4+} (calculated): 796.71 (796.71); [M+2Na+2H]^{4+} (calculated): 802.20 (802.20); [M+4Na]^{4+} (calculated): 813.20 (813.19); [M+3H]^{3+} (calculated): 1054.61 (1054.61); [M+Na+2H]^{3+} (calculated): 1061.94 (1061.94); [M+H+2Na]^{3+} (calculated): 1069.27 (1069.27); [M+3Na]^{3+} (calculated): 1076.60 (1076.60).

_{2}N-CO-Lys-(PEG_{5})_{3}-Mal_{4}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1h, 6-maleimido-hexanoic acid: 1 h; gradient: 20%–55% MeOH + 0.1% FA in 5 min, R_{t} = 4.3 min; MALDI-MS (^{+} (calculated): 3779.96 (3778.20).

_{2}N-CO-Lys-(PEG_{7})_{3}-Mal_{4}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, 6-maleimido-hexanoic acid: 1 h; gradient: 20%–55% MeOH + 0.1% FA in 5 min, R_{t} = 4.8 min; ESI-MS (^{6+} (calculated): 736.93 (736.93); [M+4H+2Na]^{6+} (calculated): 740.59 (740.59); [M+3H+3Na]^{6+} (calculated): 744.43 (744.26); [M+5H]^{5+} (calculated): 879.72 (879.72); [M+4H+Na]^{5+} (calculated): 884.12 (884.11); [M+3H+2Na]^{5+} (calculated): 888.91 (888.51).

_{2}N-CO-Lys-(PEG_{11})_{3}-Mal_{4}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, 6-maleimidohexanoic acid: 1 h; gradient: 0%–100% MeCN + 0.1% TFA in 5 min, R_{t} = 3.2 min; ESI-MS (^{7+} (calculated): 804.76 (804.76); [M+6H+Na]^{7+} (calculated): 807.9 (807.9); [M+5H+2Na]^{7+} (calculated): 811.04 (811.04); [M+6H]^{6+} (calculated): 938.72 (938.72); [M+5H+Na]^{6+} (calculated): 942.39 (942.39); [M+4H+2Na]^{6+} (calculated): 946.05 (946.05).

_{2}N-CO-Lys-PEG_{1}-Mal_{8}_{t} = 2.0 min; MALDI-MS (^{+} (calculated): 3033.76 (3033.75).

_{2}N-CO-Lys-PEG_{3}-Mal_{8}_{t} = 2.0 min; ESI-MS (^{5+} (calculated): 627.97 (627.77); [M+4H]^{4+} (calculated): 784.46 (784.46); [M+3H]^{3+} (calculated): 1046.28 (1045.61).

_{2}N-CO-Lys-PEG_{5}-Mal_{8}_{t} = 2.0 min; MALDI-MS (^{+} (calculated): 3224.24 (3223.87).

_{2}N-CO-Lys-PEG_{7}-Mal_{8}_{t} = 1.9 min; MALDI-MS (^{+} (calculated): 3312.95 (3311.93).

_{2}N-CO-Lys-PEG_{11}-Mal_{8}_{t} = 3.5 min; MALDI-MS (^{+} (calculated): 3488.57 (3488.03).

_{2}N-CO-Lys-(PEG_{1})_{4}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.4 min; MALDI-MS (^{+} (calculated): 5071.36 (5064.79); [M+2H]^{2+} (calculated): 2533.63 (2532.90).

_{2}N-CO-Lys-(PEG_{3})_{4}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 3.5 min; ESI-MS (^{7+} (calculated): 943.12 (943.12); [M+Na+6H]^{7+} (calculated): 946.41 (946.26).

_{2}N-CO-Lys-(PEG_{5})_{4}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 4.1 min; MALDI-MS (^{+} (calculated): 7946.57 (7938.58); [M+2H]^{2+} (calculated): 3962.96 (3958.80).

_{2}N-CO-Lys-(PEG_{7})_{4}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 4.7 min; MALDI-MS (^{+} (calculated): 9283.40 (9275.47); [M+2H]^{2+} (calculated): 4626.83 (4619.19).

_{2}N-CO-Lys-(PEG_{3})_{3}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 20%–50% MeOH + 0.1% FA in 5 min, R_{t} = 3.7 min; MALDI-MS (^{+} (calculated): 4623.72 (4618.67); [M+2H]^{2+} (calculated): 2310.27 (2309.84).

_{2}N-CO-Lys-(PEG_{3})_{2}-Mal_{8}_{X}: 45 min, APG-2: 2 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 20%–50% MeOH + 0.1% FA in 5 min, R_{t} = 3.3 min; MALDI-MS (^{+} (calculated): 3631.08 (3630.11).

_{2}N-CO-Lys-PEG_{11}-(PEG_{3})_{3}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 3.3 min; MALDI-MS (^{+} (calculated): 6970.43 (6970.00).

_{2}N-CO-Lys-PEG_{11}-(PEG_{3})_{2}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.4 min; MALDI-MS (

_{2}N-CO-Lys-PEG_{11}-PEG_{3}-PEG_{1}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.3 min; MALDI-MS (^{+} (calculated): 4567.08 (4562.61).

_{2}N-CO-Lys-PEG_{5}-PEG_{3}-PEG_{1}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 20%–50% MeOH + 0.1% FA in 5 min, R_{t} = 3.4 min; MALDI-MS (^{+} (calculated): 4303.96 (4298.46).

_{2}N-CO-Lys-PEG_{11}-(PEG_{1})_{3}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.4 min; MALDI-MS (^{+} (calculated): 5526.38 (5519.07), [M+2H]^{2+} (calculated): 2760.11 (2760.04).

_{2}N-CO-Lys-PEG_{11}-(PEG_{1})_{2}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–45% MeOH + 0.1% FA in 5 min, R_{t} = 2.8 min; MALDI-MS (^{+} (calculated): 4361.71 (4358.48).

_{2}N-CO-Lys-PEG_{11}-(PEG_{1})_{3}-Mal_{16}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.3 min; MALDI-MS (^{+} (calculated): 8469.42 (8470.84).

_{2}N-CO-Lys-PEG_{11}-PEG_{3}-(PEG_{1})_{2}-Mal_{16}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.4 min; MALDI-MS (^{+} (calculated): 8673.61 (8674.98), [M+2H]^{2+} (calculated): 4324.72 (4318.95).

_{2}N-CO-Lys-PEG_{11}-(PEG_{3})_{2}-PEG_{1}-Mal_{16}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.5 min; MALDI-MS (^{+} (calculated): 9086.1 (9083.25), [M+2H]^{2+} (calculated): 4527.44 (4523.09).

_{2}N-CO-Lys-PEG_{11}-(PEG_{3})_{3}-Mal_{16}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.8 min; MALDI-MS (^{+} (calculated): 9914.00 (9899.80), [M+2H]^{2+} (calculated): 4938.16 (4931.36).

_{2}N-CO-Lys-PEG_{11}-PEG_{5}-PEG_{3}-PEG_{1}-Mal_{16}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% FA in 5 min, R_{t} = 2.6 min; MALDI-MS (^{+} (calculated): 9259.57 (9259.36), [M+2H]^{2+} (calculated): 4615.74 (4611.14).

_{2}N-CO-Lys(DOTA)-PEG_{11}-Mal_{1}_{t} = 5.1 min; yield: 2.9%; MALDI-MS (^{+} (calculated): 1324.60 (1324.73), [M+Na]^{+} (calculated): 1347.37 (1346.72), [M+K]^{+} (calculated): 1362.67 (1362.82).

_{2}N-CO-Lys(DOTA)-PEG_{11}-Mal_{2}_{t} = 6.0 min; yield: 11.2%; MALDI-MS (^{+} (calculated): 1688.97 (1688.95), [M+Na]^{+} (calculated): 1711.39 (1710.93).

_{2}N-CO-Lys(DOTA)-PEG_{11}-PEG_{1}-Mal_{4}_{X}: 45 min, APG-2: 2 h, 6-maleimidohexanoic acid: 1 h; gradient: 45%–55% MeOH + 0.1% TFA in 8 min, R_{t} = 5.0 min; yield: 14.1%; MALDI-MS (^{+} (calculated): 2707.90 (2707.52).

_{2}N-CO-Cys(S_{11}-PEG_{1}-Mal_{4}_{X}: 45 min, APG-2: 2 h, 6-maleimidohexanoic acid: 1 h; gradient: 45%–65% MeOH + 0.1% TFA in 8 min, R_{t} = 6.9 min; yield: 6.4%; MALDI-MS (^{+} (calculated): 2384.42 (2384.28), [M+Na]^{+} (calculated): 2406.62 (2406.27), [M+K]^{+} (calculated): 2422.40 (2422.37).

_{2}N-CO-Lys-PEG_{11}-PEG_{1}-Mal_{4}_{X}: 45 min, APG-2: 2 h, 6-maleimidohexanoic acid: 1 h; gradient: 30%–55% MeOH + 0.1% TFA in 8 min, R_{t} = 7.7 min; yield: 17.0%; MALDI-MS (^{+} (calculated): 2321.25 (2321.34).

_{2}N-CO-Lys(DOTA)-PEG_{11}-(PEG_{1})_{2}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 30%–55% MeOH + 0.1% TFA in 8 min, R_{t} = 4.8 min; yield: 10.7%; MALDI-MS (^{+} (calculated): 4748.32 (4744.66), [M+2H]^{2+} (calculated): 2372.99 (2372.83).

_{2}N-CO-Cys(S_{11}-(PEG_{1})_{2}-Mal_{8}_{X}: 45 min, APG-2: 2 h, PEG_{Y}: 1 h, APG-3: overnight, PEG_{Z}: 2 h, APG-4: overnight, 6-maleimidohexanoic acid: 2 h; gradient: 45%–65% MeOH + 0.1% TFA in 8 min, R_{t} = 6.7 min; yield: 13.0%; MALDI-MS (^{+} (calculated): 4424.95 (4421.42), [M+Na]^{+} (calculated): 4448.00 (4443.41), [M+K]^{+} (calculated): 4465.11 (4459.51).

Reaction in solution: To a solution of the respective maleimide-multimer (

_{8}-DOTA_{t} = 4.3 min; yield: 22.8%; MALDI-MS (^{+} (calculated): 6334.62 (6334.96), [M+2H]^{2+} (calculated): 3162.4 (3156.99).

_{8}-DOTA_{t} = 4.8 min; yield: 45.2%; MALDI-MS (^{+} (calculated): 7222.99 (7223.31), [M+2H]^{2+} (calculated): 3601.52 (3600.16).

_{8}-DOTA_{t} = 4.3 min; yield: 19.8%; MALDI-MS (^{+} (calculated): 9397.17 (9392.45), [M+2H]^{2+} (calculated): 4686.46 (4685.74).

_{8}-SS-tBu_{t} = 5.5 min; yield: 18.1%; MALDI-MS (^{+} (calculated): 8156.95 (8149.09), [M+2H]^{2+} (calculated): 4068.7 (4064.06).

_{4}-DOTA_{t} = 1.8 min; MALDI-MS (^{+} (calculated): 3935.32 (3935.85).

_{4}-DOTA_{t} = 1.8 min; MALDI-MS (^{+} (calculated): 3491.79 (3491.68), [M+Na]^{+} (calculated): 3514.30 (3513.66), [M+K]^{+} (calculated): 3529.59 (3529.77), [M+Na+K]^{+} (calculated): 3552.07 (3551.75), [M+2K]^{+} (calculated): 3567.82 (3567.86).

_{4}-SS-tBu_{t} = 2.2 min; MALDI-MS (^{+} (calculated): 4239.60 (4237.12).

A solution of the respective DOTA-derivatized multimer (1.5–5 nmol) in tracepur H_{2}O was added to 244–277 MBq of ^{68}Ga^{3+} in a solution obtained by fractioned elution of a ^{68}Ge/^{68}Ga generator (Obninsk type, Eckert & Ziegler, Berlin, Germany) with HCl (0.1 M, 1.2 mL) and subsequent titration to pH 3.5–4.0 by addition of sodium acetate solution (1.25 M, 120–125 μL). After reaction for 10 min at 95 °C and cooling, the reaction mixtures were analyzed by analytical radio-HPLC. The radiolabeled products [^{68}Ga]^{68}Ga]^{68}Ga]

It was shown that the synthesis of even complex dendritic scaffold structures based on acid-amide bonds on solid supports is possible using standard solid phase peptide synthesis protocols. By systematically investigating optimal building blocks and reaction parameters, the desired homogeneous and symmetric dendrons could be obtained. In addition, molecular designs deviating from the optimized structure could also be synthesized, enabling a highly modular dendron assembly on solid support. The obtained dendritic structures comprised up to 16 maleimide functionalities and could subsequently be efficiently reacted with structurally variable thiol-bearing bioactive molecules via click chemistry. Furthermore, some dendritic scaffolds were derivatized on solid support with the chelator DOTA and finally successfully radiolabeled with ^{68}Ga after the multimerization reactions. This shows the applicability of the presented technique to the synthesis of multivalent radiotracers which can be used for molecular imaging purposes with PET.

The authors acknowledge financial support by the

G.F. performed the practical experiments and contributed to manuscript preparation, B.W. contributed to study design and C.W. contributed to practical experiments, set up the study design and prepared the manuscript.

The authors declare no conflict of interest.