New Cytotoxic Derivatives by the Phosphorylation and Phosphinoylation of Diethyl α‐Amino‐α‐Aryl‐Methylphosphonates

Abstract

In order to make available new derivatives, diethyl α‐amino‐α‐aryl‐methylphosphonates were subjected to phosphorylation, phosphinoylation and even thiophosphinoylation by reaction with phosphoryl chlorides, diphenylphosphinoyl chloride, and with the mixture of diphenylchlorophosphine and elemental sulfur, respectively. The X‐ray crystal structures of the diphenylphosphinoyl and the diphenylthiophosphinoyl derivatives revealed molecular and supramolecular similarities, as well as a few differences too. An essential conformation change, along with packing differences are attributable to a change of one heteroatom: an oxygen for a sulfur in one of the P=X function. The diethyl diethylphosphoryl‐aminobenzylphosphonates showed the highest antiproliferative effects on multiple myeloma cells, while the thiophosphinoylated diethyl aminobenzylphosphonate was the most effective on pancreatic ductal adenocarcinoma cells.

In order to make available potentionally bioactive new α‐aminophosphonic derivatives, diethyl α‐amino‐α‐aryl‐methylphosphonates were phosphorylated, phosphinoylated and thiophosphinoylated with the corresponding P‐chloride. The effect of a P=O to P=S modification was evaluated by single crystal X‐ray cristallography. Among the phosphorylated and phosphinoylated aminophosphonates a few derivatives revealed moderate cytotoxic effect.

Introduction

α‐Aminophosphonates have remained in the focus due to their real and potential bioactivity, ^[1] and due to the synthetic challenges. ^[2] α‐Aminophosphonates are known to have, among others, antiviral,^3] antibacterial, ^[4] antitumor,[ ⁵ , ⁶ ] anti‐inflammatory ^[7] and antihypertensive ^[8] effects that is the consequence of their enzyme inhibitory properties. ^[4] The synthesis of α‐aminophosphonates involves the three‐component condensation of an aldehyde/ketone, a primary/secondary amine and a >P(O)H reagent that, in most cases, is a dialkyl phosphite.[ ² , ⁹ , ¹⁰ , ¹¹ , ¹² ] The possible intermediate of the condensation under discussion maybe an imine formed from the oxo‐compound and the (primary) amine, or an α‐hydroxyphosphonate that is the adduct of the oxo‐compound and dialkyl phosphite. The phospha‐Mannich reaction may be performed under diverse conditions including the use of different metal catalysts.[ ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ ] The three‐component transformation has been the subject of many green chemical studies. ^[25] Different microwave (MW)‐assisted and solvent‐free methods were developed.[ ²⁶ , ²⁷ , ²⁸ ] Keglevich and co‐worker elaborated a MW‐assisted, solvent‐ and catalyst‐free approach ^[29] that eliminated cost and environmental burden meant by the catalyst. As a matter of fact, the mentioned intermediates may also be starting materials. Hence, the addition of a dialkyl phosphite to an imine, that is called the aza‐Pudovik‐reaction, may be a good alternative for the synthesis of α‐aminophosphonates.[ ² , ¹² ] On the other hand, it is also possible to start from an α‐hydroxyphosphonate. The latter species may undergo a nucleophilic substitution with amines. Keglevich, Mucsi et al. observed the beneficial adjacent group effect of the P=O group.[ ³⁰ , ³¹ ] Otherwise, the nucleophilic substitution on the crowded, secondary carbon atom of the hydroxyphosphonate is not easy. The modification of α‐aminophosphonates offers a good possibility to make available further derivatives that may be of potential biological activity. ^[32] Such modification may be, if an α‐aminophosphonate is reacted in a Kabachnik–Fields reaction (that is called the tandem phospha‐Mannich protocol), ^[33] or if an aminophosphonate is acylated. ^[33] It is noted that the acylaminophosphonates may also be prepared by the condensation of carboxylic amides, formaldehyde and dialkyl phosphites. ^[34]

In this article, we describe the modification of α‐aminophosphonates by phosphorylation and phosphinoylation, even thiophosphinoylation.

Results and Discussion

Preparation of Phosphorylated, Phosphinoylated and Thiophosphinoylated Aminophosphonates

In the first step, the free diethyl α‐aminophosphonates were prepared. It was found that the debenzylation of diethyl α‐benzylamino‐benzylphosphonate (1) by catalytic hydrogenation (Scheme 1) was not a clear‐cut reaction. The debenzylation was reluctant, and the purification of the complex crude mixture could not be solved. For this, we tried to utilize the method described in the literature.[ ³⁵ , ³⁶ , ³⁷ ] According to this, benzaldehyde was reacted with ammonium acetate and diethyl phosphite at 60 °C in ethanol. The imine (3 a) was decomposed by reaction with hydrochloric acid, and the free aminophosphonate (2 a) was liberated from the corresponding HCl salt (4 a) by aqueous sodium hydroxide (Scheme 2). Then, the method was extended to the preparation of aryl‐substituted α‐aminophosphonates (2 b, 2 c and 2 d).

Scheme 1.

Attempt for the debenzylation of benzylaminophosphonate 1 by catalytic hydrogenation.

Scheme 2.

Alternative route for the synthesis of aminophosphonates (2 a–d).

The substituted diethyl α‐amino‐benzylphosphonates were then taken in phosphorylation reaction using diethylphosphoryl chloride and diphenylphosphoryl chloride as the reactant and triethylamine as the base in toluene solution at 26 °C (Scheme 3). After the work‐up and purification by flash column chromatography, the phosphoryl‐aminophosphonates (5 a‐c and 6 a‐c) were obtained in 86–94 % yields. The phosphorylated aminophosphonates were characterized by ³¹P, ¹³C and ¹H NMR spectral data, as well as HRMS. The ³¹P NMR spectra of the ethoxyphosphoryl products 5 a‐c exhibited two doublets at δ_P ~7 and ~22 with a ³ J _PP coupling of ~39 Hz. The similar numbers for the phenoxyphosphoryl species 6 a‐c were δ_P ~ –2.5 and 22 along with a coupling of 43 Hz.

Scheme 3.

Phosphorylation of aminophosphonates 2 a–c.

Then, the α‐aminophosphonates 2 a‐c were subjected to phosphinoylation by reaction with diphenylphosphinoyl chloride (Scheme 4). These reactions were carried out as the phosphorylations above at room temperature. After the work‐up and purification by flash chromatography, products 7 a–d were obtained in 53–90 % yields. The phosphinoyl aminophosphonates were fully characterized. The ³¹P NMR spectrum revealed two doublets at δ_P ~22 and ~24 with a coupling of 30 Hz.

Scheme 4.

Phosphorylation of aminophosphonates 2 a–d.

The phosphonate‐phosphinic amide derivative 7 a was prepared by Polish chemists in a different way, by forming benzylidenephosphinic amide 8 in the reaction of benzaldoxime and diphenylchlorophosphine, followed by an aza‐Pudovik reaction involving the addition of diethyl phosphite onto its C=N bond (Scheme 5). ^[38]

Scheme 5.

Alternative route for the synthesis of ester‐amide 7 a. ^[38]

However, compound 7 a was only characterized by ³¹P NMR (matching our values) and IR. The claimed melting point was 134–135 °C that was 34 °C lower than ours. Their sample may not have been entirely pure.

It was a challenge to synthesize also a thiophosphinoylated derivative. This was done utilizing the three‐component reaction of aminophosphonate 2 a, diphenylchlorophosphine and elemental sulfur at 110 °C, in the presence of triethylamine in toluene under an inert atmosphere (Scheme 6). The corresponding thiophosphinoylamide – phosphonate (9) was isolated in 75 % yield and was fully characterized.

Scheme 6.

Thiophosphorylation of aminophosphonate 2 a.

X‐Ray Study

Products 7 a and 9 were subjected to single‐crystal X‐ray analysis. Figures 1 and 2 show also their supramolecular dimer structure.

Figure 1.

Crystal structure of compound 7 a with the view of the hydrogen‐bonded dimer in the crystal. DIAMOND ^[39] representation with only the asymmetric unit atoms numbered, thermal ellipsoids are drawn at 50 % probability level. Symmetry code for the non‐labelled molecule: ‐x, ‐y, ‐z.

Figure 2.

Crystal structure of compound 9 with view of the hydrogen‐bonded dimer in the crystal. DIAMOND ^[39] representation with only the asymmetric unit atoms numbered, thermal ellipsoids are drawn at a 50 % probability level. Symmetry code for the non‐labelled molecule: 1.5‐x, 0.5‐y, ‐z.

Compounds 7 a and 9 differ only in the chalcogen atom of the Ph₂P(X)‐ group: X=O for 7 a and X=S for 9. Both 7 a and 9 form hydrogen‐bonded dimers in the crystal with the participation of the NH proton and the oxygen atom of the (EtO)₂ P(O)‐group. They differ, however, in the packing of the dimeric units in the crystal, which results in different crystal systems (P 2₁/n in the case of 7 a, and C 2/c in the case of 9). A detailed comparison between the two structures reveals the influence of the chalcogen atom (X=O and S) of the Ph₂P(X) moiety in determining the crystal structure. It is worth noting that amide N atoms appear as tetrahedrons. They are slightly offset (7 a by 0.18 Å and 9 by 0.13 Å) of the planes of their P, C and H substituents. The space groups indicate that in case of compounds 7 a and 9, R,R – S,S enantiomeric supramolecular dimers form the crystal structures as racemates. For compound 7 a an S,S, while for species 9 an R,R asymmetric unit is shown in Figure 1 and Figure 2, respectively.

Simplified images showing the framework of the ring formed by the dimers can be seen in Figures 3 and 4. The bond distances of the ring structure are quite similar (Table 1). The rings formed by the dimerization are almost planar, and an inversion center is located in the middle of the ring.

Figure 3.

Simplified backbone artwork with a few selected bonds and relevant intermolecular H‐bond distances for the ring moiety of the dimer of phosphinoylamide ‐ phosphonate 7 a. Phenyl groups are indicated by their bridgehead C atoms (C2, C8 and C14) while ester groups C atoms are omitted. The short chalcogen O … H – C hydrogen bond is illustrated only by a truncated part of the respective phenyl groups.

Figure 4.

Simplified backbone illustration with a few selected bonds and relevant intermolecular H‐bond distances for the ring moiety of the dimer of species 9.

Table 1.

Bond distances with their e.s.d.s in 7 a and 9. For atomic notation convention see Figure 5.

Bond	7a	9
Oα−Pα	1.574(1), 1.572(1)	1.571(1), 1.574(1)
PαO	1.472(1)	1.472(1)
Pα−Cα	1.825(2)	1.815(2)
Cα−N	1.464(2)	1.460(2)
N−Pβ	1.651(1)	1.657(1)
Pβ=(O,S)	1.483(1)	1.950(1)
Pβ−C	1.805(2)	1.818(2)
Pβ−C	1.807(1)	1.813(2)

The two dimers formed by H‐bonding differ strongly in the orientation of the Ph₂P(X)‐group concerning the H‐bridged dimer ring. As a molecular structure backbone fit shows (Figure 5), chalcogen atoms and the opposite phenyl ring practically interchange their positions.

Figure 5.

Arbitrary atomic naming and the three relevant backbone torsion angles in 7 a and in 9. Naming is such that the anti‐periplanar position ester O atom is chosen as O _α , the acid P as P _α and so on C _α , N remains, while the chalcone portion P is as P _β , and its atom O as O _β (7 a), S as S _β (9), respectively. Phenyl appendices are designated only by Ph symbols jointly, where appropriate, and ovals representing slightly differing aromatic ring dispositions individually in 7 a and in 9. The three most relevant torsion angles are shown such that left side numbers are of 7 a while the right ones are of 9.

This is best shown by comparing the torsion angles X−P‐N−C of the two dimers; this angle is 69.9(1)° in the case of 7 a and 32.6(1)° in the case of 9, and results in a different orientation of the chalcogen atom concerning the backbone, as well as the ring formed by the dimer. A possible reason of this different orientation can be found by looking at the further weak interactions involving these two chalcogen atoms S and O. Dislike the sulfur atom of the Ph₂P(S) unit, the oxygen atom of the Ph₂P(O) group forms a short non‐classical hydrogen bond with the hydrogen atom at C5 of a neighbouring molecule. This is not surprising, because sulfur is known to be forming weaker hydrogen bridges as compared to oxygen. The O ⋅⋅H distance of 2.452(12) Å indicates, that this interaction is significant. Figure 3 shows how the centrosymmetric dimers of 7 a are fused into endless chains by the O…CH bridges. The above example underlines that the weak interactions are those, which strongly influence and, in many cases, determine the crystal structure of molecular crystals. Further weaker intermolecular contacts of X⋅⋅⋅H–C types (X=O, S) can be established (See Figures S1 and S2).

Similarities and differences between the crystal structures of the chemically close relatives 7 a and 9 can be summarized as follows:

1. Both derivatives 7 a and 9 form classic hydrogen‐bond dimers by the interaction of the O= atom of the (EtO)₂ P(O) unit and the H atom of the NH moiety (R ² ₂ (10) in Etter notation[ ⁴⁰ , ⁴¹ ]) around a symmetry center.

2. Neither the P=O of the Ph₂P(O) unit in 7 a, nor the P=S in 9 take part in classic H‐bridges.

3. Instead, they do form bifurcated acceptor=X…H–C bonds of varying strengths to 2 molecules (each from the “sides” of 2 other dimers of type I.) thus held also somewhat fixed by an R ² ₂ (12) ring forming yet another “super dimer” in species 7 a. In analogue 9, a bifurcated S acceptor is similar, but it is differing from the other that these 2 other connected molecules do not interact with each other (as in compound 7 a).

4. Amide N atoms are definitely sp ³ tetrahedrons being offset by 0.18 Å (7 a) and 0.13 Å (9) from the planes of their respective P, C and H substituent atoms, thus being in analogy with earlier foundings.[ ⁴² , ⁴³ ]

Bioactivity Study

Cell viability assays were conducted on human cancer cell lines PANC‐1 (pancreatic ductal adenocarcinoma) and U266 (multiple myeloma) at three concentrations (1, 10 and 100 μM) to assess the structure‐activity relationship and refine the selection of compounds for possible further studies. The screening results are presented in Table 2.

Table 2.

In vitro cell viability results post‐treatment on PANC‐1 and U266 cell lines after 72 h. Data represented as the mean±SD; n=3. The levels of significance are shown as follows: x: P <0.05; y: P <0.01; z: P <0.001, determined by the One‐wayANOVA test followed by Fishers LSD post hoc test.

		PANC‐1			U266
drug concentration		1 μM	10 μM	100 μM	1 μM	10 μM	100 μM
Clod		1.02±0.04	0.98±0.03	0.98±0.02	0.93±0.05	0.89±0.02z	0.88±0.04z
5a		0.9±0.04	1.05±0.03	0.93±0.01	0.77±0.02y	1.0±0.03	0.86±0.04x
5b		0.93±0.02x	1.02±0.02	0.92±0.04	0.83±0.01y	0.99±0.03	0.93±0.03 x
5c		0.85±0.18y	1.08±0.14	0.94±0.06	1.05±0.05	1.07±0.06	0.97±0.05
6a		0.73±0.01z	1.08±0.04	0.95±0.11	1.09±0.11	1.14±0.04y	0.96±0.07
6b		0.78±0.02z	1.06±0.04	0.92±0.07	1.05±0.01	1.18±0.04y	0.98±0.05
9		0.70±0.01z	1.06±0.06	1.02±0.06	1.03±0.04	1.07±0.07	1.00±0.02
7a		1.01±0.05	1.02±0.03	0.97±0.05	1.03±0.03	1.06±0.03	0.94±0.07

Clod: clodronic acid disodium salt, PANC‐1: human pancreatic ductal adenocarcinoma, U266: multiple myeloma, SD: standard deviation.

In the first case, clodronic acid disodium salt was tested as a reference molecule. ^[44] This compound belongs to the first generation of the non‐aminobisphosphonates and has been used in the treatment of bone‐related diseases, e. g., against multiple myeloma‐related bone loss. It was also found, that generally, the bisphosphonates may have anti‐proliferative activity that effect further emphasize their significance in the treatment of different malignant diseases. Interestingly, it had no remarkable antiproliferative effect on any of the cells in our long‐term experiment.

On PANC‐1 cells, among the seven investigated, newly synthesized molecules there were 6 compounds (5 a, 5 b, 5 c, 6 a, 6 b and 9) that had slightly greater antiproliferative effects than the matching clodronic acid disodium salt (clod) control. On U266 cells, two new molecules (5 a; 5 b) were found to be more effective than treatment with clod. However, these effects could only be detected at the lowest concentration (1 μM) of the compounds on both cell lines. Overall, PANC‐1 cells were less sensitive against the diethyl phosphorylamino‐benzylphosphonates (5 a‐c), than against the diethyl phosphinoylamino‐benzylphosphonate (6 a and 6 b). Despite the structural similarity of compounds 7 a and 9, as mentioned above, the thiophosphinoyl‐aminophosphonate (9) had a more prominent effect on PANC‐1 cells, than the phosphinoyl‐aminophosphonate (7 a), making species 9 the most effective on this cell line. Against the U266 cells, it was found that the phosphoryl‐aminophosphonate derivatives 5 a and 5 b were the most effective, and no further modification increased this effectiveness.

Conclusions

A few diethyl α‐amino‐α‐aryl‐methylphosphonates were modified by phosphorylation, phosphinoylation and thiophosphinoylation by reaction with phosphoryl chlorides, diphenylphosphinoyl chloride, and with the mixture of diphenylchlorophosphine and elemental sulfur, respectively, to provide new aminophosphonic derivatives with two P=O function. X‐ray analysis revealed that in the crystalline state, the phosphinoyl‐ and thiophosphinoyl aminophosphonates show identical features up to the slight difference of exchanging the oxygen atom for sulfur in function P=X, as both the conformation and the establishment of intermolecular dimeric contacts are practically the same. However, at the exchange site some conformation changes took place. The P=S function of the thiophosphinoyl moiety occupies the place of the phenyl group in the phosphinoyl counterpart, and as a consequence, the intermolecular contact strengths and their network graph sets are altered as well. Several newly synthesized aminophosphonic derivatives demonstrated greater antiproliferative effects than the clodronic acid control. Notably, the thiophosphinoyl derivative proved to be the most effective on PANC‐1 cells, while two tetraethyl phosphoryaminophosphonates exhibited the highest efficacy against U266 cells at the lowest concentration tested. Due to the lack of linearity between concentration and antiproliferative effect, further investigations are required to elucidate this phenomenon.

Experimental

General Informations

The ³¹P, ¹³C, ¹H NMR spectra were taken on a Bruker DRX‐500 or Bruker Avance‐300 spectrometer operating at 202, 126, and 500 MHz or 122, 75 and 300 MHz respectively. The ³¹ P chemical shifts are downfield relative to H₃PO₄, while the ¹³C and ¹H chemical shifts are downfield relative to TMS. The couplings are given in Hz. The exact mass measurements were performed using a Triple TOF 5600+ hybrid Quadrupole‐TOF mass spectrometer (Sciex, Singapore, Woodlands) in high‐resolution, positive electrospray mode. The melting point of product 1c was determined using a Setaram Differential Scanning Calorimetry 92 device.

General Procedure for the Synthesis of Diethyl Α‐Amino‐Benzylphosphonates 2 a–d

0.10 mol of aldehyde (benzaldehyde: 10.0 g; 4‐methylbenzaldehyde: 12.0 g; 4‐chlorobenzaldehyde: 14.0 g; 4‐methoxybenzaldehyde: 13.6 g), 0.10 mol (7.70 g) of ammonium acetate and 150 ml of ethanol were added to a 500 ml round‐bottomed flask. Then, 0.05 mol (6.44 ml) of diethyl phosphite was added to the mixture. The mixture was stirred at 60 °C for 60 hours under reflux. After completion, the excess of alcohol was removed in vacuum, then 100 ml of dichloromethane and 50 ml of water were added to the residue. Then, 10 ml of concentrated hydrochloric acid was added to adjust the pH to 1. After that, the mixture was extracted with dichloromethane, then the water phase was alkalized with 20 % aqueous NaOH solution until the pH reached 11. Then, the water phase was extracted with 2×50 ml of dichloromethane. The product was obtained from the organic phase by concentration. The organic phase was dried on Na₂SO₄ to afford products 2 a‐d (Table 3).

Table 3.

Preparative data and identification of diethyl α‐amino‐benzylphosphonates 2 a–d.

Comp.	Yield (%)	31P NMR δP (CDCl3)	δP lit.	Ref.	[M+H]+	[M+H] found ([M+H] calcd.
2 a	62	24.5	25.1	37	244	244.1109 (244.1103)
2 b	60	24.5	25.3	37	258	258.1264 (258.1259)
2 c	41	23.9	24.4	37	278	278.0723 (278.0713)
2 d	52	24.8	24.6	37	274	274.1200 (274.1208)

General Procedure for the Synthesis of Phosphorylated and Phosphinoylated α‐Aminophosphonates (5 a–c, 6 a–c and 7 a–c)

1.0 mmol of aminophosphonate (diethyl‐α‐aminobenzylphosphonate: 0.25 g; diethyl‐α‐amino‐4‐methylbenzylphosphonate 0.26 g; diethyl‐α‐amino‐4‐chlorobenzylphosphonate: 0.28 g) and 2 mL of toluene were added to a small flask with stirrer. Then, 1.0 mmol (0.14 mL) of triethylamine was added. Finally, 1.0 mmol of the phosphorylating agent (diethyl chlorophosphate: 0.14 mL; diphenyl chlorophosphate: 0.20 mL; diphenylphosphinic chloride: 0.19 mL) was added. The reaction mixture was stirred at 26 °C or at reflux for 24 hours (See Schemes 3 and 4). Then, the mixture was filtered, and the filtrate evaporated in vacuum. The crude product was purified on a 45 cm silica gel layer applying DCM:MeOH=97 : 3 as the eluent to afford the phosphorylated and phosphinoylated products.

The following compounds were thus prepared:

Diethyl Α‐Diethylphosphorylamino‐Benzylphosphonate (5 a)

Yield: 94 %, 0.35 g; ³¹P NMR (CDCl₃) δ 6.8 and 22.2 (J (P,P)=38 Hz); ¹³C NMR (CDCl₃) δ 15.7 (d, J=7.7 Hz, CH₃), 16.0 (d, J=7.4 Hz, CH₃), 16.1 (d, J=5.7 Hz, CH₃), 16.4 (d, J=6.0 Hz, CH₃), 53.0 (d, J=154.1 Hz, NCP), 62.3 (d, J=5.5 Hz, CH₂), 62.4 (d, J=5.5 Hz, CH₂), 63.0 (d, J=7.2 Hz, CH₂), 63.3 (d, J=7.1 Hz, CH₂), 127.9 (d, J=5.8 Hz, C₂’), 128.0 (d, J=3.0 Hz, C₃’), 128.4 (d, J=2.2 Hz, C₄’), 136.8 (s, C₁’); ¹H NMR (CDCl₃) δ 1.05 (t, J=7.1, 3H, CH₃), 1.11 (t, J=7.0 Hz, 3H, CH₃), 1.27 (t, J=7.1, 3H, CH₃), 1.34 (t, J=7.1 Hz, 3H, CH₃), 3.63–3.75 (m, 2H, CH₂), 3.75–3.86 (m, 1H, NH), 3.86–4.25 (m, 6H, CH₂), 4.51 (dt, J=23.1, 10.6 Hz, 1H, CH), 7.31–7.45 (m, 5H, ArH); [M+H]⁺=380; HRMS m/z: calcd for C₁₅H₂₇O₆NP₂Na [M+Na]⁺=402.1211, found: 402.1204.

Diethyl Α‐Diethylphosphorylamino‐4‐Methylbenzylphosphonate (5 b)

Yield: 92 %, 0.36 g; ³¹ P NMR (CDCl₃) δ 6.8 and 22.3 (J (P,P)=40 Hz); ¹³C NMR (CDCl₃) δ 15.7 (d, J=7.6 Hz, CH₃), 16.0 (d, J=7.5 Hz, CH₃), 16.1 (d, J=5.7 Hz, CH₃), 16.4 (d, J=6.0 Hz, CH₃), 21.1 (s, CH₃), 52.7 (d, J=154.9 Hz, NCP), 62.3 (d, J=4.8 Hz, CH₂), 62.4 (d, J=4.7 Hz, CH₂), 63.0 (d, J=7.2 Hz, CH₂), 63.2 (d, J=6.9 Hz, CH₂), 127.7 (d, J=5.9 Hz, C₂’), 129.0 (d, J=2.0 Hz, C₃’), 133.6 (s, C₄’), 137.7 (d, J=2.7 Hz, C₁’); ¹H NMR (CDCl₃) δ 1.07 (t, J=7.0, 3H, CH₃), 1.13 (t, J=7.1 Hz, 3H, CH₃), 1.28 (t, J=7.1, 3H, CH₃), 1.34 (t, J=7.1 Hz, 3H, CH₃), 2.36 (s, 3H, CH₃), 3.62–3.80 (m, 3H, CH₂ and NH), 3.97–4.25 (m, 6H, CH₂), 4.47 (dt, J=22.9, 10.6 Hz, 1H, CH), 7.12–7.38 (m, 4H, ArH); HRMS m/z: calcd for C₁₆H₃₀O₆NP₂ [M+H]⁺=394.1548, found: 395.1555.

Diethyl Α‐Diethylphosphorylamino‐4‐Chlorolbenzylphosphonate (5 c)

Yield: 90 %, 0.37 g; ³¹ P NMR (CDCl₃) δ 6.6 and 21.6 (J (P,P)=39 Hz); ¹³C NMR (CDCl₃) δ 15.8 (d, J=7.6 Hz, CH₃), 16.1 (d, J=7.3 Hz, CH₃), 16.2 (d, J=5.7 Hz, CH₃), 16.4 (d, J=5.9 Hz, CH₃), 52.5 (d, J=154.4 Hz, NCP), 62.53 (d, J=5.1 Hz, CH₂), 62.57 (d, J=5.0 Hz, CH₂), 63.2 (d, J=7.1 Hz, CH₂), 63.4 (d, J=7.0 Hz, CH₂), 128.6 (d, J=2.3 Hz, C₂’), 129.2 (d, J=5.7 Hz, C₃’), 134.0 (d, J=3.5 Hz, C₁’), 135.5 (s, C₄’); ¹H NMR (CDCl₃) δ 1.11 (t, J=7.1, 3H, CH₃), 1.17 (t, J=7.1 Hz, 3H, CH₃), 1.28 (t, J=7.0, 3H, CH₃), 1.34 (t, J=7.1 Hz, 3H, CH₃), 3.70–3.79 (m, 3H, CH₂ and NH), 3.96–4.22 (m, 6H, CH₂), 4.51 (dt, J=23, 10.6 Hz, 1H, CH), 7.33–7.40 (m, 4H, ArH); HRMS m/z: calcd for C₁₅H₂₇ClO₆NP₂ [M+H]⁺=414.1002, found: 414.1000 for the ³⁵Cl isotope.

Diethyl Α‐Diphenylphosphorylamino‐Benzylphosphonate (6 a)

Yield: 89 %, 0.42 g; mp: 103–104 °C; ³¹ P NMR (CDCl₃) δ −2.7 and 21.6 (J (P,P)=43 Hz); ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.8 Hz, CH₃), 16.4 (d, J=5.9 Hz, CH₃), 53.5 (d, J=154.7 Hz, NCP), 63.2 (d, J=7.2 Hz, CH₂), 63.5 (d, J=7.0 Hz, CH₂), 120.1 and 120.2 (J=5.0 Hz, C₂”), 124.9 (d, J=10.0 Hz, C₂’), 128.0 (d, J=5.8 Hz, C₃’), 128.1 (d, J=2.7 Hz, C₄’), 128.5 (bs, C₄”) 129.48 and 129.55 (s, C₃”), 135.9 (s, C₁’), 150.56 and 150.57 (d, J=6,9 Hz, C₁”); ¹H NMR (CDCl₃) δ 1.06 (t, J=7.0 Hz, 3H, CH₃), 1.24 (t, J=7.1 Hz, 3H, CH₃), 3.56–3.65 (m, 1H, NH), 3.83–4.25 (m, 4H, CH₂), 4.78 (dt, J=22.8, 10.5 Hz, 1H, CH), 7.01–7.30 (m, 15H, ArH); HRMS m/z: calcd for C₂₃H₂₈O₆NP₂ [M+H]⁺=476.1392, found: 476.1391.

Diethyl Α‐Diphenylphosphorylamino‐4‐Methylbenzylphosphonate (6 b)

Yield: 86 %, 0.42 g; mp: 140–141 °C; ³¹ P NMR (CDCl₃) δ −2.5 and 21.7 (J (P,P) =43 Hz); ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.5 Hz, CH₃), 16.4 (d, J=5.8 Hz, CH₃), 21.1 (s, CH₃), 53.2 (d, J=155.6 Hz, NCP), 63.1 (d, J=7.2 Hz, CH₂), 63.4 (d, J=6.9 Hz, CH₂), 120.2 and 120.3 (J=5.0 Hz, C₂”), 124.8 (d, J=5.7 Hz, C₂’), 128.0 (d, J=5.9 Hz, C₃’), 129.1 (bs, C₄”), 129.4 and 129.5 (s, C₃”), 133.0 (s, C₄’), 137.7 (d, J=2,8 Hz, C₁’), 150.7 (d, J=6.7 Hz, C₁”); ¹H NMR (CDCl₃) δ 1.08 (t, J=7.0 Hz, 3H, CH₃), 1.24 (t, J=7.1 Hz, 3H, CH₃), 2.35 (s, 3H, CH₃), 3.57–3.69 (m, 1H, NH), 3.83–4.20 (m, 4H, CH₂), 4.74 (dt, J=22.5, 10.4 Hz, 1H, CH), 7.02–7.28 (m, 14H, ArH); HRMS m/z: calcd for C₂₄H₃₀O₆NP₂ [M+H]⁺=490.1548, found: 490.1546.

Diethyl Α‐Diphenylphosphoryllamino‐4‐Chlorobenzylphosphonate (6 c)

Yield: 91 %, 0.46 g; mp: 139–140 °C; ³¹ P NMR (CDCl₃) δ −2.3 and 20.9 (J (P,P)=43 Hz); ¹³C NMR (CDCl₃) δ 16.2 (d, J=5.5 Hz, CH₃), 16.4 (d, J=5.7 Hz, CH₃), 52.9 (d, J=155.9 Hz, NCP), 63.38 (d, J=6.3 Hz, CH₂), 63.42 (d, J=7.1 Hz, CH₂), 120.1 and 120.2 (J=4.9 Hz, C₂”), 124.9 (d, J=7.4 Hz, C₂’), 128.5 (bs, C₄”), 129.4 (d, J=5.7 Hz, C₃’), 129.5 (s, C₃”), 133.9 (bs, C₁’), 134.8 (s, C₄’), 150.5 and 150.6 (J=6.8 Hz, C₁”); ¹H NMR (CDCl₃) δ 1.10 (t, J=7.1 Hz, 3H, CH₃), 1.24 (t, J=7.0 Hz, 3H, CH₃), 3.64–3.73 (m, 1H, NH), 3.86–4.25 (m, 4H, CH₂), 4.75 (dt, J=22.9, 10.4 Hz, 1H, CH), 7.05–7.32 (m, 14H, ArH); HRMS m/z: calcd for C₂₃H₂₇ClO₆NP₂ [M+H]⁺=510.1002, found: 510.1012 for the ³⁵Cl isotope.

Diethyl Α‐Diphenylphosphinoylamino‐Benzylphosphonate (7 a)

Yield: 90 %, 0.47 g; mp: 168–169 °C; ³¹ P NMR (CDCl₃) δ 22.3 and 24.3 (J (P,P)=30 Hz), δ lit ^[38] 21.7 and 23.9 (J (P,P)=29.3 Hz); ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.6 Hz, CH₃), 16.5 (d, J=5.9 Hz, CH₃), 52.0 (d, J=153.4 Hz, NCP), 63.2 (d, J=7.4 Hz, CH₂), 63.3 (d, J=8.0 Hz, CH₂), 127.8 (s, C₄’), 128.06 (bs [overlapped], C₂’), 128.10 (d, J ~11 Hz) and 128.4 (d, J ~14 Hz) C₃”, 128.32 (bs [overlapped], C₃’), 131.8 (s) and 132.0 (s [overlapped]) C₄”, 132.0 (d, J=10.0 Hz) and 132.3 (d, J=9.9 Hz) C₂”, 131.9 (d, J~133 Hz) and 132.3 (d, J~128 Hz) C₁”, 136.9 (s, C₁’); ¹H NMR (CDCl₃) δ 1.06 (t, J=7.1 Hz, CH₃), 1.37 (t, J=7.0 Hz, CH₃), 3.60–3.71 (m, 1H, CH₂), 3.88–3.98 (m, 1H, CH₂), 4.23–4.27 (m, 2H, CH₂), 4.36–4.48 (m, 1H, NH), 4.62 (dt, J=22.5, 10.6 Hz, CH), 7.25–7.96 (m, 15H, ArH; HRMS m/z: calcd for C₂₃H₂₈O₄NP₂ [M+H]⁺=444.1494, found: 444.1479.

Diethyl Α‐Diphenylphosphinoylamino‐4‐Methylbenzylphosphonate (7 b)

Yield: 66 %, 0.60 g; dense oil; ³¹ P NMR (CDCl₃) δ 22.5 and 22.2 (J (P,P)=29.0 Hz); ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.5 Hz, CH₃), 16.5 (d, J=5.9 Hz, CH₃), 21.1 (s, C–CH₃), 51.8 (d, J=154.2 Hz, NCP), 63.15 (d, J=5.5 Hz, CH₂), 63.20 (d, J=7.3 Hz, CH₂), 127.9 (d, J=5.9 Hz, C₂’), 128.1 (d, J=12.7 Hz) and 128.4 (d, J=12.7 Hz) C₃”, 129.0 (s, C₃’), 131.7 (bs) and 131.9 (bs) C₄”, 132.1 (d, J=9.9 Hz) and 132.3 (d, J=10.0 Hz) C₂”, 132.0 (d, J=127.9 Hz) and 132.9 (d, J=130.1 Hz) C₁”, 133.8 (bs, C₁’), 137.4 (bs, C₄’); ¹H NMR (CDCl₃) δ 1.08 (t, J=7.1 Hz, CH₃), 1.37 (t, J=7.0 Hz, CH₃), 2.34 (s, C–CH₃), 3.59–3.69 (m, 1H, CH₂), 3.85–3.94 (m, 1H, CH₂), 4.13–4.27 (m, 3H, CH₂ + NH), 4.50 (dt, J ₁=22.3 Hz, J ₂ =10.8 Hz, 1H, CH), 7.07–7.91 (m, 14H, ArH); HRMS m/z: calcd for C₂₄H₃₀O₄NP₂ [M+H]⁺=458.1650, found: 458.1652.

Diethyl Α‐Diphenylphosphinoylamino‐4‐Chlorobenzylphosphonate (7 c)

Yield: 53 %, 0.50 g; dense oil; ³¹ P NMR (CDCl₃) δ 21.8 and 24.5 (J (P,P)=28.7 Hz); ¹³C NMR (CDCl₃) δ 16.2 (d, J=5.8 Hz, CH₃), 16.5 (d, J=5.9 Hz, CH₃), 51.5 (d, J=154.0 Hz, NCP), 63.4 (d, J=7.2 Hz, CH₂), 63.5 (d, J=7.3 Hz, CH₂), 128.3 (d, J=12.9 Hz) and 128.53 (d, J=12.8 Hz) C₃”, 128.54 (d, J=2.2 Hz, C₃’), 129.4 (d, J=5.9 Hz, C₂’), 132.1 (d, J=11.1 Hz) and 132.0 (d, J=10.3 Hz) C₂”, 132.1 (d, J=128.4 Hz) and 132.2 (d, J=128.2 Hz) C₁”, ~132.2 (C₄”) overlapped by 133.8 (d, J=3.4 Hz, C₁’), 135.5 (d, J=1.8 Hz, C₄’); ¹H NMR (CDCl₃) δ 1.07 (t, J=7.1 Hz, CH₃), 1.35 (t, J=7.0 Hz, CH₃), 3.58–3.67 (m, 1H, CH₂), 3.78–3.88 (m, 1H, CH₂), 4.15–4.23 (m, 2H, CH₂), 4.54 (dt, J ₁=22.6 Hz, J ₂=10.8 Hz, 1H, CH), 4.37–4.47 (m, 1H, NH), 7.24–7.90 (m, 14H, ArH); HRMS m/z: calcd for C₂₃H₂₇ClO₄NP₂ [M+H]⁺=478.1104, found: 478.1082 for the ³⁵Cl isotope.

Diethyl Α‐Diphenylphosphinoylamino‐4‐Methoxybenzylphosphonate (7 d)

Yield: 65 %, 0.60 g; dense oil; ³¹ P NMR (CDCl₃) δ 22.5 and 24.3 (J (P,P)=29.3 Hz) ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.7 Hz, CH₃), 16.4 (d, J=5.9 Hz, CH₃), 51.4 (d, J=164.0 Hz, NCP), 55.2 (s, OCH₃), 63.05 (d, J=5.1 Hz, CH₂), 63.14 (d, J=5.1 Hz, CH₂), 113.7 (s, C₃’), 128.1 (d, J=12.8 Hz) and 128.3 (d, J=12.8 Hz) C₃”, 128.9 (d, J=1.9 Hz, C₁’), 129.3 (d, J=6.0 Hz, C₂’), 131.6 (d, J=2.6 Hz) and 131.8 (d, J=2.8 Hz) C₄”, 132.0 (d, J=9.8 Hz) and 132.2 (d, J=10.0 Hz) C₂”, 132.1 (d, J=127.0 Hz) and 132.4 (d, J=127.4 Hz) C₁”, 159.1 (d, J=2.6 Hz, C₄’); ¹H NMR (CDCl₃) δ 0.99 (t, J=7.5 Hz, CH₃), 1.28 (t, J=7.5 Hz, CH₃), 3.50–3.59 (m, 1H, CH₂), 3.72 (s, CH₃O), 3.75–3.84 (m, 1H, CH₂), 4.09–4.19 (m, 3H, CH₂ + NH), 4.41 (dt, J ₁= 20.0, J ₂=11.7 Hz, 1H, CH), 6.72–7.82 (m, 14H, ArH); HRMS m/z: calcd for C₂₄H₃₀O₅NP₂ [M+H]⁺=474.1599, found: 474.1577.

Diethyl‐Α‐Diphenylthiophosphinoylamino‐Benzylphosphonate (9)

2.0 mmol (0.52 g) of diethyl‐α‐aminobenzylphosphonate, 2.0 mmol (0.28 mL) of triethylamine and 6 mL of toluene were added to a small round‐bottomed flask. Then, 3.0 mmol (0.1 g) of sulfur was added to the mixture. Finally, 3.0 mmol (0.55 mL) of chlorodiphenylphosphine in 2 mL of toluene was added. The reaction mixture was stirred at 110 °C for 4 hours under reflux. The mixture was filtered, and then the filtrate evaporated in vacuum under nitrogen. The crude product was purified on a 35 cm silica gel layer applying DCM:MeOH=97 : 3 as the eluent. The purified product crystallized spontaneously, forming brown, needle‐like crystals. Yield: 75 %, 0.55 g; mp: 180–181 °C; ³¹ P NMR (CDCl₃) δ 22.5 and 60.9 (J (P,P)=38 Hz); ¹³C NMR (CDCl₃) δ 16.1 (d, J=5.6 Hz, CH₃), 16.4 (d, J=6.2 Hz, CH₃), 52.0 (d, J=155.4 Hz, NCP), 63.0 (d, J=7.3 Hz, CH₂), 63.1 (d, J=7.1 Hz, CH₂), 127.7 (d, J=2.5 Hz, C₄’), partially overlapped with 127.8 (d, J=13.3 Hz) and 128.2 (d, J=11.7 Hz) C₃”, 128.1 (s, C₃’), 128.8 (d, J=6.2 Hz, C₂’), 131.1 (d, J=3.2 Hz) and 131.38 (d, J=3.0 Hz) C₄”, partially overlapped with 131.40 (d, J=11.4 Hz) and 131.6 (d, J=11.7 Hz) C₂”, 135.1 (d, J=99.3 Hz) and 135.2 (d, J=104.0 Hz) C₁”, 136.8 (s, C₁’); ¹H NMR (CDCl₃) δ 1.01 (t, J=7.0 Hz, CH₃), 1.30 (t, J=7.1 Hz, CH₃), 3.54–3.67 (m, 1H, CH₂), 3.80–3.99 (m, 2H, CH₂ and NH), 4.09–4.24 (m, 2H, CH₂), 5.0 (dt, J=25.0, 10.0 Hz, CH), 7.18–7.93 (m, 15H, ArH); HRMS m/z: calcd for C₂₃H₂₈O₃NP₂S [M+H]⁺=460.1265, found: 460.1246.

Single Crystal X‐Ray Diffraction Studies

Single crystals of compounds 7 a and 9, suitable for X‐ray diffraction, were obtained by slow evaporation of acetone solutions. The crystals were introduced into perfluorinated oil and a suitable single crystal was carefully mounted on the top of a thin glass wire. Data collection was performed with an Oxford Xcalibur 3 diffractometer equipped with a Spellman generator (50 kV, 40 mA) and a Kappa CCD detector, operating with Mo‐K_α radiation (λ=0.71071 Å).

Data collection and data reduction were performed with the CrysAlisPro software. ^[45] Absorption correction using the multiscan method ^[45] was applied. The structures were solved with SHELXS‐97, ^[46] and finally checked using PLATON. ^[47] Details for data collection and structure refinement are summarized in Table 4.

Table 4.

Details for X‐ray data collection and structure refinement for compounds 7 a and 9.

	7 a	9
Empirical formula	C23H27NO4P2	C23H27NO3P2S
Formula mass	443.39	459.45
T[K]	123(2)	123(2)
Crystal size [mm]	0.30×0.15×0.10	0.30×0.15×0.08
Crystal description	colorless block	colorless block
Crystal system	monoclinic	monoclinic
Space group	P21/n	C2/c
a [Å]	10.5962(2)	23.1106(5)
b [Å]	15.5568(3)	12.8347(3)
c [Å]	13.7020(3)	16.1067(3)
α [°]	90.0	90.0
β [°]	94.853(2)	103.555(2)
γ [°]	90.0	90.0
V [Å3]	2250.58(8)	4644.45(18)
Z	4	8
ρcalcd. [Mg m−3]	1.309	1.314
μ [mm–1]	0.222	0.302
F(000)	936	1936
Θ range [°]	2.86–25.24	2.60–25.24
Index ranges	−14≤h≤14	−29≤h≤29
	−20≤k≤20	−16≤k≤16
	−18≤l≤18	−20≤l≤20
Reflns. collected	40068	36291
Reflns. obsd.	4745	4384
Reflns. unique	5571 (Rint=0.0463)	5127 (Rint=0.0378)
R 1, wR 2 (2σ data)	0.0398, 0.1044	0.0346, 0.0866
R 1, wR 2 (all data)	0.0484, 0.1113	0.0433, 0.0913
GOOF on F 2	1.055	1.044
Peak/hole [e Å–3]	0.66 /‐0.43	0.51/‐0.27

Supplementary crystallographic data are available for compounds 7 a and for 9, under Deposition Numbers 2391623 and 2391624, respectively. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe http://www.ccdc.cam.ac.uk/structures.

Cytotoxic Experimental

Cell Culturing

The two investigated cell lines were a human myeloma (U266) and a human pancreatic ductal adenocarcinoma (PANC‐1) purchased from the European Collection of Authenticated Cell Cultures (ECACC, Salisbury, UK).

The PANC‐1 cell line has an adherent characteristic (87092802 ECACC), while the U266 cells (85051003 ECACC) grow in suspension. The PANC‐1 was cultured in DMEM medium (Sigma Ltd., St. Louis, MO, USA), and the U266 cells were cultured in RPMI 1640 (Sigma Ltd., St. Louis, MO, USA) medium. In both cases, the medium was supplemented with 10 % fetal bovine serum (Invitrogen Corporation,New York, NY, USA), 1 % L‐glutamine (Invitrogen Corporation, New York, NY, USA) and 1 % penicillin/streptomycin (Invitrogen Corporation, New York, NY, USA).

Cell Viability Assays

The investigated molecules were solubilized in dimethyl sulfoxide (DMSO; AppliChem GmbH, Darmstadt, Germany) (stock solutions: 10⁻¹ M). The v/v% of DMSO was kept under 1. Clodronic acid disodium salt was solubilized in sterile distilled water (stock solution: 10^–1 M). The stock solutions were kept at −80 °C and in each case, fresh solutions were prepared for each experiment.

A non‐invasive impedimetric method (xCELLigence system) was employed to test the viability of PANC‐1 cells post‐treatment. The adherent cells were seeded at a concentration of 105 cells/mL in a 96‐well plate (E‐Plate 96 PET; ACEA Biosciences, San Diego, CA, USA) whose bottom is covered by gold microelectrodes that can detect changes of impedance which correlates with cell number, allowing the measurement of the compounds’ antiproliferative effects. Prior to cell seeding, a baseline was established using a cell‐free medium. After overnight incubation, the cells were treated with the investigated molecules (final concentration: 1, 10 and 100 μM) as well as the corresponding controls (medium and DMSO). Data were recorded using the RTCA 2.0 software (Real Time Cell Analyzer; ACEA Biosciences, San Diego, CA, USA).

Due to the suspension characteristics of U266 cells, no test in the xCELLigence system could be conducted. Consequently, the CellTiter‐Glo Luminescent Cell Viability Assay (Promega, Madison, WI, USA) was used instead. Cells were seeded at a density of 105 cells/mL in a white‐walled 96‐well plate (Thermo Scientific, Waltham, MA, USA). After overnight incubation, the cells were treated with the investigated molecules (final concentration: 1, 10 and 100 μM) as well as the corresponding controls (medium and DMSO). Following 72 hours of incubation, CellTiter‐Glo Reagent was added to each well, and the luminescence signal was measured using the Fluoroskan FL Microplate Fluorometer and Luminometer (Thermo Scientific, Waltham, MA, USA).

Both experiments were performed in triplicate, and results were normalized to the untreated medium control and expressed as mean±SD.

Supporting Information Summary

³¹P, ¹³C and ¹H NMR spectra of the compounds (5 a–c, 6 a–c, 7 a–d and 9) prepared together with additional details of the X‐ray analysis.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Bircher M., Karaghiosoff K., Czugler M., Takács A., Kőhidai L., Drahos L., Keglevich G., Chem. Eur. J. 2025, 31, e202500370. 10.1002/chem.202500370

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.