Toxicity, bioactivity, release of H2S in vivo and pharmaco-kinetics of H2S-donors with thiophosphamide structure

Article abstract of DOI:10.1016/j.ejmech.2019.05.017

H₂S donors are substitutes of H₂S with various biological activities like inhibiting the inflammatory response and protecting myocardial cells from injury. In order to confirm whether the H₂S donors have drug-like properti

Full text of DOI:10.1016/j.ejmech.2019.05.017

Accepted Manuscript
Toxicity, bioactivity, release of H S in vivo and pharmaco-kinetics of H S-donors with
2
2
thiophosphamide structure
Jinlong Zhang, Qiuping Zhang, Yanni Wang, Jili Li, Zhongjie Bai, Quanyi Zhao, Dian
He, Zhen Wang, Jingke Zhang, Yonglin Chen
PII:
S0223-5234(19)30426-X
DOI:
https://doi.org/10.1016/j.ejmech.2019.05.017
EJMECH 11327
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 27 February 2019
Revised Date: 5 May 2019
Accepted Date: 6 May 2019
Please cite this article as: J. Zhang, Q. Zhang, Y. Wang, J. Li, Z. Bai, Q. Zhao, D. He, Z. Wang, J.
Zhang, Y. Chen, Toxicity, bioactivity, release of H S in vivo and pharmaco-kinetics of H S-donors
2
2
with thiophosphamide structure, European Journal of Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.ejmech.2019.05.017.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Toxicity, bioactivity, release of H2S in vivo and pharmaco-
kinetics of H2S-donors with thiophosphamide structure
Jinlong Zhang1, Qiuping Zhang1, Yanni Wang1, Jili Li1, Zhongjie Bai1, Quanyi Zhao *1,
Dian He1, Zhen Wang1, Jingke Zhang2, Yonglin Chen3
(1 Institute of Medicinal Chemistry, School of Pharmacy of Lanzhou University; 2 GLP Lab
Centre, School of Basic Medicine, Lanzhou University; 3 The First Affiliated Hospital of
Lanzhou University, Lanzhou, 730000, China)
Abstract: H2S donors are substitutes of H2S with various biological activities like
inhibiting the inflammatory response and protecting myocardial cells from injury. In
order to confirm whether the H2S donors have drug-like properties, two series
thiophosphamide H2S donors were evaluated including toxicity, bioactivity and
pharmacokinetic properties in vivo and in vitro. The following results were obtained.
Firstly, all the compounds released H2S under measuring condition; with the increase
of pH value, the H2S release rate of all the compounds decreased and the amount
reduced, but pH value had little effect on the maximum release of H2S. Secondly, in
the organs and tissues of rats, the compounds released H2S in the same way as in PBS.
In plasma, compound 1 reached the Cmax after administration 55 min, and no
compound 1 was detected after 12 h; for compound 18, the Cmax reached only after
administration 100 min, and after 6 h, compound 18 was not detected; in organs and
tissues, the H2S-release rates were different from those in PBS, but the mechanism of
H2S release was the same. Thirdly, in the test of toxicity, all the compounds displayed
low toxicities to 5 cancer cells and W138 cell lines; compounds 1 and 18 had slight
effect on the physiological tissue and function of rat liver at low concentration; the
compounds had almost no effect on the hatching rate, survival rate of zebrafish
embryos, and the spontaneous movement of zebrafish embryos at below 0.5 µM, but
when they were over 1 µM, the compounds displayed inhibitory effect in the manner
of concentration dependence. Fourthly, in the course of anti-inflammatory test, all the
tested compounds significantly reduced the level of TNF-α and increased the level of
IL-10; when they were 100 µM, the levels of IL-10 were three times as high as those
in the control group. Among them, compounds 10 and 18 displayed stronger activities
than the others. In addition, the compounds protected H9c2 cells from injure and
improved myocardial injury through anti-oxidation pathway. In summary, the
compounds have druglike properties due to low toxicity, better activity and good
pharmacokinetic property. Therefore, they have potential to be as candidates to
investigate further.
Corresponding author, Email: zhaoqy@lzu.edu.cn

ACCEPTED MANUSCRIPT
Keywords: H2S donors; thiophosphamide; toxicities; anti-inflammation; myocardial
protection; pharmacokinetic
Introduction
H2S is considered as the third gas signal molecule after CO and NO.1-6 Many
tests confirmed H2S displayed various biological activity, for example, inhibiting the
inflammatory response.7,8 Endogenous H2S was produced by enzymatic reaction using
cysteine and homocysteine as substrates.9-15 In the past decade, a lot of testing results
show H2S plays an important role in physiology and pathology process, especially in
the cardiovascular system and nervous system.16-22 Besides that, H2S also induced
cancer cells apoptosis by activating map kinase and cysteinase-3. 23 All indicates H2S
has a big potential to be candidate drugs.24,25
However, H2S is a rotten-egg gas, and inconvenient to use in clinic due to
uncontrollable dose. Consequently, using H2S donors taking the place of H2S gas to
use in clinic is a practical way. H2S donors have inorganic sulfide salts and organic
26
H2S donors. The salts like NaSH, Na2S and CaS, release H2S almost immediately,
and possibly lead to the toxic injury of tissue cells because the concentration of H2S is
too high. 27 The organic H2S donors may overcome these faults to get controllable
dose and targeting tissues and organs. At present, they involve mainly thioamides,
disulfide and compounds containing polysulfide atoms. 28-34
Among the organic H2S donors, thiophosphamide derivative is a kind of H2S
donors with low cytotoxicity. The structures of thiophosphoramide donors include
cyclic phosphoramide structure and chain phosphoramide structure.35-36 They all
released H2S in the process of test, and exhibited various biological activities, such as
35
anti-inflammation, myocardial protection.
However, although there are many
reports about H2S donors, there is still no any report on the kinetic properties of H2S
donors in vivo. Whether the H2S donors can be used in clinic as medicines, there are
many determinants. Among them, the toxicity and the pharmacokinetic properties of
compound in vivo are crucial. In addition, H2S donor is only a carrier of H2S, they do
not take effect until H2S released from it arrives at the effective concentration. In vivo,
the microenvironments (such as pH) of organs and tissues maybe affect the H2S
releasing rates. In order to understand deeply and comprehensively the properties of
H2S donor compounds in vivo, it is very necessary to investigate the H2S-release in
organs and tissues.
Therefore, in this paper, based on the synthesis of two series of thiophosphamide
H2S donors, we evaluated the toxicity and bioactivity using several models, and
investigated their pharmacokinetic properties in vivo. By which, we hope to obtain
some information about whether the H2S donors can be as candidates to be further
study, and also provide theoretical basis for the clinical application.

ACCEPTED MANUSCRIPT
Results and discussion
1. Synthesis and characterization of compounds:
Compounds 1-9 were synthesized by the reaction of phenylthiophosphoryl
dichloride with different amino alcohols, diols or diamines in the presence of
triethylamine. The product was purified by column chromatography to obtain white
granular solid. 36Compounds 10-18 also took phenyl thiophosphonyl dichloride as the
starting material, reacting with 3-hydroxypropionitrile and amino acid methyl ester to
obtain ester intermediates; then the esters hydrolyzed with LiOH to form lithium salt,
following freeze-dried to form white granular products.35 Nine amino acid methyl
esters were involved (glycine, phenylalanine, valine, alanine, proline, leucine,
isoleucine, tryptophan and methionine). All the compounds were characterized by 1H
NMR, 13C NMR, ESI-MS and IR.
Scheme 1 The synthetic route and structures of compounds 1-9

ACCEPTED MANUSCRIPT
Scheme 2 The synthetic route and structures of compounds 10-18
The spectra of all the compounds are consistent with the expected ones. IR data
1
and other main characteristics are shown in Table 1. In the H NMR spectra of
compounds 1-9, the chemical shifts of hydrogen atoms in benzene ring appeared in
the range of 7.88-7.28 ppm. For all the compounds containing O atom (all but
compounds 4 and 6), the signals of protons on OCH2 appeared in the range of 3.40-
4.49 ppm. For compounds containing NH group (compounds 4, 6 and 9), the chemical
shifts of hydrogen in NH are in the range of 3.15-3.44 ppm. In the hydrogen spectra
of compounds 10-18, the chemical shifts of hydrogen in benzene ring were in the
range of 7.81-7.07 ppm; meanwhile, the chemical shifts of hydrogen in NH appeared
in the range of 3.42-3.86 ppm and those in NCH2 in the range of 3.52-3.82 ppm.

ACCEPTED MANUSCRIPT
Table 1 The main characteristics (solubility, IR absorption) of the compounds
Compds
solubility
IR characteristic absorption
（cm-1）
stability
trait
3274m (NH),1438s (P=S),
1120s (C-O)
white granular
solid
1
stable in dry
air at r.t ;
3265m (NH), 1436s (P=S),
1121s (C-O)
1440s (P=S), 1125s (C-O)
light yellow
oil
white oil solid
2
3
4
5
6
7
insoluble in
water;
3252m (NH), 1436s (P=S)
1439s (P=S), 1118s (C-O)
3317m (NH), 1437s (P=S)
white oil solid
white oil solid
easily
soluble in
organic
solvents such
as DMSO,
dichlorometh
ane,acetone,
and so on
do not decay
for long
periods of
time
yellow
granular solid
white oil solid
3448m (NH), 1437s (P=S),
1115s (C-O)
3418m (NH), 1437s (P=S),
1120s (C-O)
white oil solid
8
3437m (NH), 1439s (P=S),
1161s (C-O)
white granular
solid
9
3370m (NH), 1593s (COO),
1436.9 (s, P=S)
3476s (NH), 1636s (COO),
1411s (P=S)
3472m (NH), 1634s (COO),
1418s (P=S)
3428m (NH), 1641s (COO),
1416 (s, P=S)
3434m (NH), 1641s (COO),
1410 (s, P=S)
3294m (NH), 1638s (COO),
1437s (P=S)
3420m (NH), 1638s (COO),
1422s (P=S)
3404m (NH), 1636s (COO),
1420s (P=S)
10
11
12
13
14
15
16
17
18
easily soluble
in water,
methanol and
DMSO;
easily
deliquescent
and tend to
deteriorate in
humid air
white granular
solid
insoluble in
dichlorometh
ane, ethanol,
acetone
3431m (NH), 1638s (COO),
1418s (P=S)
2. H2S release ability of the compound and its influencing factors
Among the methods used to detect H2S, methylene blue method was used
generally, but it needs to be carried out under strong acid condition, and the promotors
such as cystine often lead to a large deviation in the measurement results.37 In contrast,
the fluorescence probe method is obviously superior to methylene blue method. The
fluorescence probe is an important tool for selective and sensitive detection of H2S.
The fluorescence probe based on sulfonyl azide can react with H2S in a few minutes,

ACCEPTED MANUSCRIPT
showing a higher fluorescence change and excellent linear correlation. So the H2S
released from the donor can be measured quantitatively by the fluorescence value of
the probe. 38
In the testing of process, we chose fluorescence probe based on sulfonyl azide to
detect the H2S released from the compounds. The results show all the compounds
released H2S at room temperature, and most of the H2S release curves are in
accordance with the first order kinetic equation. All the data is shown in SI-Table 1.
Seen from the data, the first series compounds released H2S far more slowly than
the second series, and accordingly the H2S release half lives were longer than those of
the second; moreover, the maximum amounts of H2S-release of the first series
compounds were also lower than those of the second series under the same condition.
For compounds 1-9, their maximum releases of H2S were all in range from 25 to 30
µM, which indicates that the size of cycle-type structure and the function groups like
ester and amide had little effect on the release of H2S. Compounds 1-9 were H2S
releasers with medium release rates, and their half-lives were about 30 min. For
compounds 10-18, except for 14, the maximum release of H2S of each compound was
more than 100 µΜ and in the range of 100-120 µM. Compound 14 released H2S a
small amount and only 4.03 µM at 22 . This is possibly because the influence of the
rigid structure of tetrahydropyrrole gave rise to the carboxyl anion far away the N-P
double bond not to take part in the intramolecular substitution reaction. The half-life
of compound 14 was about 10 min while the others were about 3.5 min, which
indicates the second series of compounds were fast H2S releasers.
In the body, the pH value of each organ and tissue varies, for example, the pH of
gastric juice is about 1.8, and the pH of small intestine is about 8.0 when the body
temperature is 37 . When the H2S donor enters the body, is the H2S-release affected
by the pH value and temperature? In order to simulate as much as possible the real
situation of the release of H2S in vivo, the half-life and maximum release of
compound were measured under different pH and temperature conditions. All data is
also shown in Table 2 and SI-Table 1.
Table 2 H 2 S Cmax released from the compounds at 150 µM
H2S Cmax(µM)
22℃
pH 7.4
37℃
pH 7.4
Com
-pds
pH 1.8
25.0
pH 5.0
24.9
pH 8.0
25.6
pH 1.8
34.8
pH 5.0
35.8
pH 8.0
35.1
25.1
29.9
24.4
27.2
26.5
31.1
27.2
27.7
1
2
3
4
30.0
29.9
29.7
35.6
37.0
35.4
24.8
25.7
25.7
38.4
37.4
35.7
27.0
27.2
27.1
35.7
35.5
35.7

ACCEPTED MANUSCRIPT
28.4
27.9
27.7
27.5
35.7
35.7
29.1
35.7
5
28.5
28.4
28.4
28.3
35.7
35.4
29.9
34.8
6
27.1
27.1
26.9
27.1
36.2
36.8
28.2
33.7
7
25.5
25.5
25.4
28.1
36.7
36.6
26.4
35.4
8
28.3
28.5
27.8
29.2
35.4
34.5
30.0
34.7
9
101.4
121.5
121.4
104.2
4.0
99.4
99.5
99.4
102.6
128.8
125.2
109.7
4.2
103.0
128.6
125.9
106.5
4.2
102.2
128.3
125.3
107.7
4.2
101.4
127.0
124.2
105.8
4.3
10
11
12
13
14
15
16
17
18
121.0
121.5
102.9
4.2
121.5
121.5
102.4
4.0
120.3
121.5
102.7
4.1
113.1
118.6
114.3
118.4
111.8
117.0
112.6
117.2
111.5
117.3
112.3
117.3
111.4
117.0
112.6
118.4
118.6
122.9
120.0
124.4
118.3
123. 0
119.8
124.4
118.2
122.7
119.9
124.3
115.5
122.4
118.1
123.5
Fig. 1 H2S concentration change with time when compound 1,7,10,14,17 and 18 at 150 µM, and
probe 200 µM: A for 22°C; B for 37°C.

ACCEPTED MANUSCRIPT
Fig.2 The effect of pH on H2S release of the compounds at 150 µM, probe 200 µM. (A) for
compound 1; (B) for compound 7; (C) for compound 17; (D) for compound 18.
Seen from Fig.1, compared with room temperature, when the temperature was at
37 , the maximum release of H2S of compounds 1 and 7 increased slightly, but their
half-lives did not change significantly. For the second series, the maximum releases of
H2S of compounds 10, 17 and 18 were also slightly increased, and their half lives of
release of H2S were also shortened, while compound 14 did not change. In general, at
37 , the release rates of all the compounds increased and the release amount
increased. Meanwhile, the pH value of the solution also affected the H2S release. As
shown by Fig.2, the release half lives and maximum releases of compounds 1 and 7
changed little, but compounds 17 and 18, the smaller pH value, the faster the H2S
releasing rate. With the increase of pH value, the release of H2S became slow and the
amount became less, but pH had no effect on the maximum release of H2S. Of course,
except pH value and temperature, the substances with nucleophilic group also
promote H2S release, such as cysteine 39-41 and glutathione
.
3. H2S release ability of the compounds in organs and tissues
3. 1. The mechanism of the compound releasing H2S
The H2S release ability of the compound is closely related to its structure. The
first series compounds (such as compound 1) are more stable, and the H2S released
slowly in PBS; the mechanism of H2S-release was proposed as showing in Fig.3 (a).
The second series compounds (e.g compound 18) released H2S more quickly, and the
process of H2S release in PBS was considered as showing in Fig.3 (b). 35 Whether the
H2S-release process as we expected? Is there any difference between in vivo and in

ACCEPTED MANUSCRIPT
vitro? In vivo, whether the products formed by the compound after releasing H2S are
the same as those done in PBS?
（a）
H2S
O
P
S
P
S
P
SH
P
O
O
O
O
OH
N
H2O
H O
N
N
H
N
2
H
H
H
I (a)
II (a)
III （ a）
(b)
O
P
S
O
P
O
O
P
C6H5
S
C6H5
P
O
Li
C6H5
SH
Li
C6H5
P
O
Li
O
C6H5
H2O
H
HN
HN
R
HN
HN
HN
R
R
R
R
H2S
O
O
O
O
O
O
O
Li
O
O
Li
Li
II (b)
III (b)
I (b)
Fig. 3 Mechanism of H2S- release of the compounds as H2S donors: (a) for compound 1, (b)
for compound 18.
To illustrate this problem, we used LC-MS to detect the products after the
compound releasing H2S in PBS and serum. Seen from Fig.4 and SI-Fig. 2(a), in PBS,
the retention time (Rt) of compound 1 was 9.649 min, and the characteristic peaks in
MS appeared at m/z 214.0 ([M+H]+) and m/z 236.0 ([M+Na]+). The Rt of the product
was about 8.652 min, and the characteristic peaks appeared at m/z 198.00 ([M+H]+)
and m/z 220.00 ([M+Na]+). In serum (Fig.4 and SI-Fig.2(b) ), the Rt of compound 1
was 9.647 min, and the characteristic peaks also appeared at m/z 214.0 ([M+H]+) and
m/z 236.0 ([M+Na]+); the Rt of the product was 8.851 min, and the characteristic
peaks also appeared at m/z 198.00 ([M+H]+ and m/z 220.00 ([M+Na]+), which
indicated the product was the same as that in PBS. The product was identified as
compound III-(a) by comparison of known substance. Therefore, we infer that
compound 1 released H2S in serum and in PBS in the same manner, and is consistent
with our expectation.
There are many various components in serum, whether they have an effect on the
release rate of H2S. To answer this question, the concentrations of compound in both
PBS and serum were detected every 12 h using LC-MS in the process of test. Seen
from Fig.4, in PBS, the concentration of compound 1 decreased gradually and the
product III-(a) increased gradually with the prolongation of time in 36 h; in contrast,
in serum, it was much lower than that in PBS buffer of pH 7.4 at the same time point.
This is because compound 1 had a certain protein binding rate. At the same time, in
serum, the concentration of the product III-(a) was higher than that in PBS. This
indicates compound 1 released H2S faster in serum than in PBS.

ACCEPTED MANUSCRIPT
Fig. 4 LC diagrams of compound 1 over time in PBS (A) and serum (B)
Compound 18 was detected every 30 min because of the short half-life of H2S
release. Seen in SI-Fig.4, the Rt of compound 18 was about 8.08 min, and the
characteristic peak appeared at m/z 318.10 ([M+H]+). Compound 18 released H2S
easily in serum, and III-(b) was detected at the beginning of test. The content of
compound III-(b) increased quickly with the prolongation of time (SI-Fig.5b).
Compared with PBS buffer (SI-Fig.5a), the content of compound 18 was much lower,
which shows it had a higher protein binding rate. After H2S release, III-(b) was the
main product which Rt was about 7.27 min, and characteristic peak at m/z 302.10
([M+H]+). It was reported that compound 18 released H2S by forming cyclic
compound II-(b), 35 but we did not detect compound II-(b) in the process of test. This
suggests the formation of cyclic compound II-(b) is not the driving force of H2S
release from compound 18. However, we found compound 18 also released H2S in
methanol solution (containing little water) during the treatment of serum samples, and
cyclic compound II-(b) was the main product except H2S.

ACCEPTED MANUSCRIPT
3.2. Process of H2S release of the compounds in liver, heart and kidney
The compounds displayed the pharmacological activity of H2S in the organs and
tissues like myocardial tissue, which demonstrates them having been distributed to
target organs and tissues and released H2S. When the compound enters the body, it
may be disposed by the enzymes in various tissue cells. In the liver, the compounds
possibly do not release H2S due to being oxidized. In cardiomyocyte, is there any
toxical substance produced after H2S release from the compound? In kidney, does the
compound also release H2S in the same manner as in other organs? In order to figure
out this problem, we used the liver, heart and kidney homogenate of WISTAR rats to
simulate the process of H2S release in vivo. During the process of test, the method of
H2S measure was the same as above. The results showed all the tested compounds
released H2S in various tissue homogenates.
In liver homogenate, compound 1 decreased with time. As shown in Fig.5(B),
after incubation for 36 h, most of compound 1 still remained intact; the primary
metabolite increased gradually, which the retention time was 8.63 min, and the
characteristic peaks appeared at m/z 198.0 ([M+H]+) and 220.0 ([M+Na]+). It is
obvious that the primary metabolism is compound III-(b).
In myocardial homogenate, compound 1 released H2S was also slow. As shown
in Fig.6, except for the metabolite III-(b), two new products appeared, and their
retention times were 11.9 min and 12.4 min, respectively. The characteristic peaks in
appeared at m/z 449.0 and m/z 662.1. This indicates that compound 1 transformed
into new products in myocardial tissue after releasing H2S; however, these two new
compounds also appeared in kidney tissue homogenate (SI-Fig.2(c) and 3).
Fig. 5 LC-MS diagram of compound 1 in liver homogenate. A for 0 h and B for 36 h

ACCEPTED MANUSCRIPT
Fig. 6 LC-MS diagram of compound 1 in heart homogenate. A for 0 h; B for 36 h
For compound 18, the release rate of H2S in liver homogenate was still fast. In
the LC diagram, with exception of the original compound 18 (Rt 8.3 min, m/z 318.1
[M+H]+) and III-(b) (Rt 7.9 min, m/z 302.1[M+H]+), there were two products, which
their retention times were 10.2 min and 10.6 min; moreover, their contents changed
obviously with time. These two substances were also found in serum, heart, and
kidney tissues (SI-Fig.5(c, d, e)), and their concentrations also decreased with time.
Their m/q values are higher, so we thought they maybe were small peptide conjugates
of compound 18. These conjugates prolonged the action time of H2S. Possibly, this is
the reason why the half life of H2S release was short in PBS, but it had obvious
myocardial protection in vivo. Therefore, it was in liver, heart and kidney tissues that
compound 18 released H2S in the same way as in PBS, but the release rates and the
amount of metabolites are slightly different.
Seen from the above results, these two series compounds released H2S in both
PBS and various organs and tissues in the same way. The protein binding rate of
compound 1 in serum, while compound 18 had high protein binding rate, which
making H2S release slowly. In different tissues, the rates of release H2S of these two
compounds in vitro were different due to the effect of plasma protein, but the
mechanism of H2S release was the same.
4. Pharmacokinetics of the compounds as H2S donors
Male WISTAR rats were chosen as animal model, and injected intraperitoneally
with compound 1 or compound 18 of 200 mg/kg. The blank group was injected with
the solution without containing compound 1 or 18. The blood was taken from canthus
eye of the rats every set time point. After treatment, the compound was analyzed by
LC-MS.
As shown in Fig.7, compound 1 was absorbed quickly into blood. After 50 min
post administration, the plasma concentration of compound 1 reached maximum, and
then it decreased sharply over time, which indicates the distribution rate of compound
1 into organs and tissues was also fast. After 6 h, the concentration of compound 1
was very low and tended to the detection limit. After 12 h, compound 1 was not
detected in the plasma of rats, which suggests compound 1 was metabolized into other
substances or excreted completely. For compound 18, the plasma concentration
reached maximum post 100 min administration. After 6 h, compound 18 was not
detected in the plasma, indicating that it was metabolized into other substances or

ACCEPTED MANUSCRIPT
excreted out of the body. In comparison, compound 1 was absorbed faster than
compound 18, while compound 1 was eliminated slower than compound 18, possibly
because the polarity of compound 1 is so small that it easily passes through the cell
membranes to be absorbed. Compound 18 was easily eliminated because it has a big
polarity.
For the same compound, there is a big difference of property between in vivo and
in vitro. This is because the compound was distributed, metabolized and eliminated
gradually in vivo. This explains why compound 1 still can be detected in vitro after 36
h while cannot be detected in vivo after 12 h.
Fig.7 Blood concentration curves of compounds vs time, 1 (left) and 18 (right). Each data
was measured three times and its average value was taken.
5. Toxicity of the compounds as H2S donors
5.1 Cytotoxicity of the compounds as H2S donors
In order to evaluate the cytotoxicity of the compounds to cells, six cell lines
(HeLa, A549, HePG2, MCF-7, HT-29 and W138 cell line ) were chosen to evaluate
the inhibitory effect of the compounds on the growth of the cells, 5-FU as positive
control. The results showed all the compounds had little inhibitory effect on the
growth of six cell lines compared with 5-FU; some of the IC50 values were above 800
µM (Table 3). Among all the first series compounds, compound 2 displayed a slight
effect against HT-29 cells, with IC50 539.1 µM; compound 5 had a better inhibitory
effect on HeLa cells than the others.
The second series of compounds also displayed low toxicities to the cells, possibly
because their structures contain amino acid fragments. Among the second series,
compound 16 displayed the strongest inhibitory effect on the growth of A549 cells
with an IC50 of 236.7 µM. The anti-proliferative activity of compound 17 to HePG2
cells was higher than the others, and its IC50 was 285.9 µM. That indicates the toxicity
of compound 17 was much lower than that of 5-FU. Compound 12 displayed higher
anti-proliferation against MCF-7 cells than the others, but its IC50 was still very large.
All in all, all 18 compounds had a low inhibitory activity on the growth of HeLa,

ACCEPTED MANUSCRIPT
A549, HePG2, MCF-7, HT-29 and W138 cell lines; Moreover, the compounds
displayed no selectivity to cancer cells and normal cells.
Table 3 IC50 values of all compounds against cells proliferation
IC50(µM)
Compds
HeLa
A549
HePG2
MCF-7
HT-29
W138
1
2
826.3±13.4
886.1± 11.8
628.7±9.6
752.0±16.2
1054.9±10.8 720.5±12.8
929.9±8.7
621.7±16.9
775.0±11.7
612.0±15.4
800.6±17.4
804.3±15.7
656.9±10.1
531.0±10.2
596.7±8.8
797.5±14.8
852.9±13.1
850.1±13.0
748.5±9.8
712.8±13.4
920.5±17.7
979.9±9.3
890.5±8.0
875.7±8.2
662.8±9.8
542.4±13.2
832.0±15.5
874.4±11.6
643.3±9.9
601.3±15.3
671.4±12.2
837.6±12.1
136.5±14.1
539.1±16.0
991.1±16.7
600.4±14.0
882.1±17.0
780.5±12.3
635.5±13.5
1005.4±8.8
653.9±9.5
758.8±16.4
652.8±10.7
380.6±17.4
381.0±9.5
957.8±14.8
502.9±16.5
481.1±8.3
990.2±10.5
508.5±13.1
689.5±11.3
545.7±9.1
351.4±12.2
871.6±14.7
438.5±16.3
658.2±15.8
546.2±17.5
723.6±13.5
201.9±8.7
3
1278.8±16.0 569.3±11.6
4
793.3±9.0
678.7±14.6
967.5±8.5
984.3±10.0
875.6±10.9
678.6±15.3
641.4±11.2
543.8±11.3
5
6
7
8
1429.8±15.9 886.4±12.7
9
945.3±13.8
913.1±12.9
711.9±13.5
809.4±10.8
888.4±17.3
838.7±13.4
835.2±11.0
972.1±15.3
919.1±17.8
712.3±13.1
120.4±12.6
675.4±16.0
377.7±14.0
236.7±15.9
294.5±8.9
798.1±17.0
770.7±16.9
671.1±12.5
689.6±17.0
858.2±11.2
815.5±12.1
414.7±11.1
603.9±16.6
285.9±8.9
10
11
12
13
14
15
16
17
18
5-FU
912.4±13.3
542.1±15.2
602.3±13.6
689.3±13.9
932.2±18.0
756.4±11.5
782.4±17.7
826.2±14.9
741.6±9.0
345.6±10.4
518.7±10.7
273.3±14.4
236.3±18.0
308.8±10.8
433.3±12.7
104.5±11.9
315.1±13.0
112.1±13.4
146.2±16.2
Note: Values represent mean ± SD from at least three independent experiments.
5.2 Toxicity of the compound after successive administration
In the set dose range, the single dose of mice did not show obvious toxicity, so
we selected the rats to study the toxicity of the compounds by multiple doses. The
liposolubility of the first series of compounds is large, and their lipids and water
partition coefficients are between 2.1 and 3.1, indicating they are easy to permeate the
membrane and are easily absorbed; the second series of molecules are the lithium salt
structure, which is not easy to absorb and can only be administered by intraperitoneal
injection. The tested compounds were injected intraperitoneally with 0.01 mmol•kg-1
per dose each day. After 14 days, the rats were killed; then the blood, liver, kidney
and other main organs were collected, and examined at the same time.
The results of routine blood test (SI-Table 2) showed, after the rats were
administrated the compound successively for 14 days, the number of white blood cells
decreased compared with the control (5.82×109/L); and the numbers were 3.23 and

ACCEPTED MANUSCRIPT
4.01×109/L for compound 1 and 18, respectively. The platelet numbers were 628 and
530×109/L, which was also lower than that of the control group (857×109/L).
However, the hemoglobin content and erythrocyte number were increased. This
indicates that the negative impacts of the compound on the blood system are in
controllable range.
Fig.8 Morphology of liver (A and B), kidney (C and D) and spleen (E and F) of the rats after
successive administration for 14 days. A, C and E for compound 1; B, D and F for compound 18
In order to understand the degree of damage caused by the compound to the
physiological structure of organs of the rats, we observed the liver, kidney and spleen
tissues under the microscopic after H-E staining. The results (Fig.8) showed, in the
control group, the hepatic lobules were well structured, the hepatocytes were arranged
neatly, and there was no obvious gap between the cells. In test group, the hepatic
lobules were still intact after 14 days of administration (Fig.8 (A and B)), but center
venous and the hepatic sinuses were slight congestion. The hepatocytes were swollen
and marked gaps were observed. In the kidney section, the glomerular was intact in
the control group, the space of proximal convoluted tubules was normal; in the test
group, the glomeruli were damaged slightly after administration (Fig.8 (C and D)):
the cells of proximal convoluted tubules were swollen and marked gaps were
observed. For splenic sections, compared with the control, some nucleus became
larger, and there was slight inflammation (Fig.8 (E and F)). These results suggest the
compounds will damage to the liver, kidney and spleen of rats after long-term
administration. Therefore, the compounds are suitable for short-term successive
administration.

ACCEPTED MANUSCRIPT
6. Effect of the compound on the development of zebrafish
6.1 Effect of the compound on hatching rate and survival rate of zebrafish
embryos
In order to assess the development toxicity of the compounds, we selected
zebrafish embryos as model animal. The hatching rate and survival rate of zebrafish
embryos were measured at 24, 48, 72, 96 and 120 hpf. The results show: when the
compound was below 0.5 µM, the compound had almost no effect on survival rate of
zebrafish embryos after 120 hpf compared with the control (Fig.9 (A and B)), and
also had less effect on hatching rate of zebrafish (Fig.9(C)). For each tested
compound, the embryo survival rate decreased with the increase of the compound
concentration. When compound 1 or 18 was up to 20 µM, at 120 hpf, both the embryo
survival rate and hatching rate decreased to 50% or so; moreover, the longer the
incubating with the compound under the same condition, the lower both of them.
The incubation period of normal zebrafish embryos ranges from 48 to 72 hpf. In
the incubation process, the compounds inhibited the key enzymes to promote the
embryonic development, and then led to the embryonic of zebrafish stopping the
growth.[42] Seen from Fig.9(C), for compounds 1 and 18, the hatching rate of
zebrafish embryos was almost the same as the survival rate of embryos: at low
concentrations (0.5 and 1.0 µM), the hatching rate of zebrafish embryos was not
significantly affected by the compounds compared with the control; however, it
decreased sharply when the compounds were over 5.0 µM. All these results indicate it
is the compounds with low concentration that will be safe to the development of
zebrafish embryos.

ACCEPTED MANUSCRIPT
Fig.9 A(for compound 1) and B (for compound 18) is the effect of the concentration of
compound on the embryo survival of zebrafish; C is the effect of the concentration of
compound on the hatching rate of zebrafish embryosat 48 hpf; D is the effect of the
concentration of compound on the spontaneous movement of zebrafish at 24hpf; E is the effect
of the concentration of compound on the deformity rate of zebrafish embryos; F is the effect of
the concentration of compound on the total swimming distance of juvenile zebrafish; n = 100. *
compared with the control (DMF), * p < 0.05, * * p < 0.01. The data shown represent mean
value ±SD for three independent experiments. DMF as solvent control, and compared with
blank control, the diluted DMF (1: 1000) had no effect on zebrafish experiment.
6.2 The effect of compound on the spontaneous movement of zebrafish embryos
In order to evaluate the neurotoxicity of compound, the spontaneous motion
frequency of zebrafish embryos was measured in 1min at 24 hpf. The results are
shown in Fig.9 (D). As a whole, the spontaneous motion frequency of zebrafish

ACCEPTED MANUSCRIPT
embryos decreased gradually with the increase of compound concentration. When the
concentration was below 0.5 µM, compound 18 showed no effect on the spontaneous
motion frequency of zebrafish embryos compared with the control (17 times / min).
However, at higher concentration, this effect was significant. Compound 1 of 5.0 µM
decreased the motion frequency to 12 times /min while compound 18 decreased it to
13 times /min. This suggests the both compounds had slight neurotoxicity to zebrafish
embryos at higher concentration, and this effect was concentration dependent.
6.3 Teratogenicity of the compound on zebrafish
At 24 hpf, zebrafish embryos were treated with the compound solution of 0.5 µM,
and then observed under microscope at 96 hpf. Seen from Fig.9 (E), the compounds
had no effect on the development of zebrafish embryos at below 0.5 µM. When
compound 1 was in the range of 1-20 µM, the embryo malformation rate was about
2%; but as for compound 18, the embryo malformation rate increased with the
increase of concentration under the same condition, and when it reached 20 µM, the
embryo malformation rate was about 5%. This suggests the teratogenicity was
concentration-dependent. The embryos incubated with compound 1 (or 18) appeared
pericardial edema, spinal curvature and yolk cysts (Fig.10). Therefore, we conclude
that the compounds caused embryo malformation at over 1 µM, and they are safe at
below 0.5 µM.
Fig.10 The morphology of zebrafish embryos after incubated with the compound; A for normal
zebrafish; B for compound 1; C for compound 18. Red arrows indicate malformation. PE stands
for pericardial edema; VC for yolk sac; BS for spinal curvature.
6.4 Effects of the compounds on juvenile movement of zebrafish
In order to make out whether the compounds have effect on the development of
nervous system of zebrafish, we assessed the effect of compound on the motor ability
of juvenile zebrafish at 144 hpf. The results show, under visible light irradiation, the
total swimming distance of juvenile decreased with the increase of compound
concentration. Seen from Fig.9 (F), the compounds with higher concentration
significantly inhibited juvenile movement, but when they were below at 1.0 µM, there
was less difference in total swimming distance compared with the control. When
compound 1 was 10 µM, the total swimming distance of juvenile zebrafish in 5 min
was 257 mm; but for compound 18, it was 216 mm under the same condition.
Therefore, we conclude the tested compounds have a concentration dependent toxicity
to the development of the nervous system of juvenile zebrafish. Compared with

ACCEPTED MANUSCRIPT
compound 1, compound 18 displayed more serious toxicity to the development of
nervous system at the same concentration.
7. The anti-inflammatory activity of the compounds
We chose to use macrophage RAW264.7 as model to test and verify the anti-
inflammatory activities of the compounds. In order to obtain more accurate results
and exclude the effects of cell death on the experimental results, the survival rate of
cells was measured by MTT method after treated with tested compounds at different
concentrations (20 µM and 100 µM) for 24 h in the presence or absence of LPS (1
µg/ml). The results showed that the tested compounds had little toxicity to
macrophages (Fig.11), and the survival rates of macrophages reached 70-110%. This
indicates the anti-inflammatory results are authentic when the tested compounds are
used in the range of 20-100 µM.
Fig.11 The effect of the compound on macrophage RAW 264.7 survival rate after treated for 24
h. In each experiment, four groups of parallel experiments were set, and the data were
represented by Mean ±SD.
NO is one of the inflammatory mediators, and excessive NO up-regulates the
level of inducible nitric oxide synthase (iNOS), which is a key step in the
development of inflammation.43-44 NO can be easily oxidized to nitrite by superoxide
ions in the body. H2S can not only quench peroxynitrite, but also react with NO to
form nitromercaptan.45-47 Therefore, the anti-inflammatory activities of the
compounds as H2S donors were evaluated by measuring the amount of nitrite.

ACCEPTED MANUSCRIPT
LPS can activate iNOS to generate more NO. To make the effect more obvious,
the macrophages activated by LPS were used to evaluate the anti-inflammatory
activities of the compounds. The test results showed all the tested compounds reduced
the nitrite levels in RAW 264.7 cells in a concentration-dependent manner, as shown
in Fig.12. Among the tested compounds, compound 18 exhibited the strongest activity
and decreased by about 50% nitrite level at 100 µM compared with the control. This
is possibly because compound 18 has the largest H2S release amount among four
compounds.
Fig.12 The effect of the concentration of compound on nitrite in macrophages RAW264.7 after
treated for 24 h, A for compound 1, B for 8, C for 10 and D for 18. Compared to untreated cells,
#p < 0.05, ## p < 0.01, compared to LPS treated cells, * p < 0.05, ** p < 0.01.In each experiment,
four groups of parallel experiments were set, and the data were represented by mean ±SD.

ACCEPTED MANUSCRIPT
Fig.13 The effect of the concentration of compound on the level of TNF-α in macrophages after
treated for 24 h; A for compound 1, B for 8, C for 10 and D for 18. Compared to untreated cells,
#p < 0.05, ## p < 0.01, compared to LPS treated cells, * p < 0.05, ** p < 0.01.In each experiment,
four groups of parallel experiments were set, and the data were represented by mean ±SD.
TNF-α is a tumor necrosis factor, which promotes T cells to produce various
inflammatory factors, further promoting the occurrence of inflammatory reaction. As
shown in Fig.13, all the tested compounds significantly reduced the level of TNF-α;
among them, compounds 10 and 18 displayed stronger activities than compounds 1
and 8 when their concentrations were 20 µM. The level of TNF-α decreased to 20-30%
of the control group. In contrast, compounds 1 and 8 reduced TNF-α level to 30-40%
of the control group under the same conditions. IL-10 is an anti-inflammatory factor.
The level of IL-10 in cell also is one of indexes of inflammation. Seen in Fig.14, all
the tested compounds increased the level of IL-10 in the cells. Among four
compounds, compounds 10 and 18 exhibited better activities than compounds 1 and 8.
When compounds 10 and 18 were 20 µM, the levels of IL-10 were two times as high
as those in the control group. Therefore, we conclude compounds 10 and 18 have
better anti-inflammatory activity than compounds 1 and 8.

ACCEPTED MANUSCRIPT
Fig.14 The effect of the concentration of compound on IL-10 expression in macrophages after
treated for 24 h, A for compound 1, B for 8, C for 10 and D for 18. Compared to untreated cells,
#p < 0.05, ## p < 0.01, compared to LPS treated cells, * p < 0.05, ** p < 0.01. In each experiment,
four groups of parallel experiments were set, and the data were represented by mean ±SD.
In conclusion, all the tested compounds showed good anti-inflammatory activity.
They all reduced nitrite content in macrophage RAW264.7, and compound 18
displayed the highest activity among them, which closely related to the maximum
release of H2S. The compounds reduced the level of TNF-α and increased the level of
IL-10 in a dose-dependent manner, thus inhibiting inflammatory mediators and
exerting anti-inflammatory activity. This further indicates that H2S play an important
role in the anti-inflammatory process.
8. Myocardial protection of compounds
In order to avoid the test error due to H9c2 cells death caused by the compound,
we first use MTT method to measure the survival rate of H9c2 cells after treated with
the compound. Seen from SI-Fig.6, the four compounds have no significant effect on
the proliferation of H9c2 cells. When the concentration reached 100 µM, the survival
rate still exceeded 90%. This indicates that the concentration below 100 µM is
reliable and effective during the process of testing. Consequently, the tested
compounds all were used below 100 µM in the process of following test.

ACCEPTED MANUSCRIPT
Fig.15 The effect of the concentration of compound on the H9c2 cell survival rate after treatment
with H2O2 for 8h: A for 50µM; B: for 200 µM H2O2; C for 400 µM H2O2. Each bar represents the
average of six independent experiments ±sd, compared with model group, * p < 0.05; ** p < 0.01 .
In order to get the protective effect of the compound on myocardial injury more
remarkable, the H9c2 cells were treated with different concentration H2O2 for 8 h.48
Seen from Fig.15, after treatment with H2O2 of 50 µM, 200 µM and 400 µM for 8 h,
the survival rates of H9c2 cells were 75.6%, 45.6% and 20.3%, respectively; however,
after the H9c2 cells were incubated with tested compounds, the survival rates
increased. When H2O2 was 50 µM, compound 1 of 50 µM made the cell survival rate
increase by 17.1%; however, when H2O2 was 200 and 400 µM, the cell survival rates
were only increased by 13.9% and 5.1%, respectively. The more serious the cell
damage, the lower the protective effect of the compound. When H2O2 was 400 µM,
the damage of the cells is too serious to repair. In the range of 10-100µM, the activity
of compound 1 was not concentration-dependent. When H2O2 was 50 µM and 200
µM, the optimal dose of the compound was 50 µM, and when the concentration was
increased to 100 µM, the protective effect of the cells began to decrease.
As for compound 18, when H2O2 was 50 µM and 200 µM, compound 18 of 50
µM increased the survival rates of the cells by 25.5% and 30.1%, respectively.
However, after the cells treated with 400 µM H2O2 for 8 h, the cell survival rate still
decreased even though compound 18 arrived at up to 100 µM, even was lower than
that of H2O2 control group. It is clear that H2O2 of 400 µM is beyond the safe
concentration range of the cells, which is consistent with the results of toxicity test.
Compared with compound 1, compound 18 displayed higher activity than compound
1. This is because compound 18 released more H2S than compound 1.
After the H9C2 cells treated with H2O2 and compound were stained with Gimsa
dye for 30 min, microscope was used to observe morphology of H9c2 cells.

ACCEPTED MANUSCRIPT
Fig.16 Morphology of H9c2 cells after treatment with H2O2-compound (50 µM), blank group
(A, B and c); H2O2 group (D, G and J); H2O2-compound 1 group (E, H and K); H2O2-compound
18 group (F, I and L). Cells were cultured in a 24-well plate and treated for 24 h. From left to
right, 50 µM, 200 µM and 400 µM H2O2 were used for 8h, respectively.
As shown in Fig. 16, in the control group, H9c2 cells were long fusiform, firmly
adhered and not abnormal (Fig. 16 (A, B and C)), but in H2O2 group, the cells were
irregular, not closely arranged, and some cells became round (Fig. 16 (D, G and J)).
With the increase of oxidative damage of H2O2, more and more deformable cells were
damaged and the damage was more and more serious. In the presence of compounds 1
or 18, some of the injured cells caused by H2O2 returned to their original state.
However, when H2O2 was 400 µM, most injured cells did not restore to their original
state, which indicates that the damage of the cells was too serious and the compound
did not display the protective role. The results correspond to the above experiments.
Fig.17 The effect of the concentration of compound on SOD and MDA levels in H9c2 cells: A:
SOD; b: MDA. The data showed an average ±SD, n = 6. * p < 0.05 * p < 0.01.

ACCEPTED MANUSCRIPT
Generally, the oxidative stress reaction happened in cells leads to cell damage
and even cell death. MDA and SOD levels in cells are two indexes in oxidative stress
reaction. 49-50 In order to seek the reason why the compound protected myocardial
cells from oxidative injury, we measured the levels of SOD and MDA levels in H9c2
cells after treatment with H2O2-compound. The results show cell damage caused by
H2O2 resulted in a decrease in the level of SOD in the cells, but in the presence of the
tested compound, the SOD level increased distinctly (Fig.17 (A)). Meanwhile,
treatment of H9c2 cells with H2O2 resulted in an increase in MDA, but MDA levels
significantly decreased after incubating with the compound (Fig.17 (B)). Moreover,
the antioxidant capacity was concentration dependent. Among the tested compounds,
compound 18 showed better antioxidant activity than the others. All suggest the
compounds protected myocardial cells from injure possibly by through increasing
SOD level and decreasing MDA level.
Conclusion
H2S is a gas signal molecule and has various biological activities like inhibiting
the inflammatory response and inducing cancer cells apoptosis. H2S plays an
important role in the cardiovascular system and nervous system. H2S donors are
substitutes of H2S. Many tests indicate H2S donors have a big potential to be used in
clinic.
In order to confirm whether the H2S donors have drug-like properties, two series
thiophosphamide H2S donors were synthesized here. The drug-like properties involve
several aspects, including bioactivity, toxicity, pharmacokinetic properties and so on.
Because H2S donors reported focused on synthesis and bioactivity, there is no any
study about pharmacokinetic properties in vivo. Consequently, the pharmacokinetic
property of the compounds is our main issue to be solved. We found pH value was
different, the H2S-release rates and amounts of the compounds were different, which
suggests the compounds release H2S differently in different organs and tissues in vivo.
Importantly, each compound released H2S in both PBS and organ and tissues in the
same way. In rat models, the pharmacokinetic properties of the compounds vary with
their structures: compound 1 had a certain protein binding rate in plasma; the plasma
concentration reached maximum after administration 55 min and no compound 1 was
detected in the plasma after 12 h; whereas compound 18, the protein binding rate was
high; the maximum plasma concentration reached after administration 100 min, and it
was metabolized or eliminated completely 6 h later. In addition, in organs and tissues
in vivo, the rates of release H2S of the compounds were different from those in vitro,
but we found the mechanism of H2S release was the same.
In toxicity aspect, all the compounds displayed lower cytotoxicity, and displayed
almost no toxicity to the mice in the process of single dose administration. But after
successive administration for 14d, the number of white blood cells in the blood of rats
reduced, and the liver and kidney of rats slightly damaged. In addition, when the
compounds were below 0.5 µM, they had almost no effect on the hatching rate and
survival rate of zebrafish embryos, the spontaneous movement of zebrafish embryos,
and it did not caused embryo malformation. But when the concentration was over 1

ACCEPTED MANUSCRIPT
µM, the compounds displayed significant inhibitory effect in the manner of
concentration dependence. This suggests the compounds are safe for short-term
period of application at low dose. Finally, we testified the compounds not only
displayed better anti-inflammation activity, but also protected H9c2 cells from injure
through increasing SOD level and decreasing MDA level, such as 18.
In summary, based on the above, we draw a conclusion the compounds have
druglike properties, and they have potential to be as candidates to investigate further.
However, we have to see that the H2S donor as a drug will be used in clinic, there's a
long way to go. Since H2S as an endogenous gas signaling molecule, has a variety of
biological activities, such as anti-inflammatory, myocardial protection, anti-tumor and
so on, what patients are fit to H2S donor to deal with? This may be a question we will
discuss in the future.
Experimental
Reagents and instruments
All reactions were carried under nitrogen atmosphere. Solvents for reactions
were degassed and distilled from the proper drying agents. Column chromatography
was carried out using 200–300 mesh silica gel; RAW264.7 macrophages lines, HeLa,
A549, HePG2, MCF-7, HT-29 and WI38 cell lines were from cell resources Center
for Shanghai Life Science Institute of Chinese Academy of Sciences (China),
zebrafish from Center for Shanghai Life Science Institute of Chinese Academy of
Sciences (China).
IR spectra were recorded on a Nicolet NEXUS 360 spectrophotometer, and
NMR spectra on a Bruker AM-400 MHz spectrometer. A Lambda 25 UV–Visible
spectrophotometer and a Maxis-4G TOF Mass spectrometer (ESI) were used.
Fluorescence intensity was recorded using an RF-5301PC fluorescence
spectrophotometer. Pharmacokinetics of the compounds was studied using the Agilent
1260 ininfinity II LC and Agilent infinity Lab LC/MSD.
Preparation of compounds 1~9
Compound 1:
A mixture solution of 3-aminopropanol (0.52 mL, 6.77 mmol) and triethylamine
(2.0 mL, 9.0×10-5 mol) in CH2Cl2 was added dropwise to a solution of phenylthiono
phosphonic dichloride (1mL, 6.77 mmol) in CH2Cl2 with stirring. The mixture was
stirred for 12 h in room temperature, the precipitate was filtered off, the filtrate was
washed with cold water (3x20 mL), dried with Na2SO 4, and the solvent was removed
in vacuum. The crude product was chromatographed on a silica gel (PE/EA 3:1), and
a white granular solid was obtained. Yield: 37.4%. IR (KBr, cm-1): 3274 (NH),1438
1
(P=S), 1120 ( C-O); H NMR (400 MHz, CDCl3) δ (ppm): 7.79 (m, 2H, Ar-H), 7.42

ACCEPTED MANUSCRIPT
(m, 3H, Ar-H), 4.42 (m, 1H, OCH2), 4.02 (m, 1H, OCH2), 3.38(br, 1H, NH), 3.11 (m,
13
2H, NCH2), 2.00 (m, 1H, CH2), 1.58 (m, 1H, CH2); C NMR (100 MHz, CDCl3) δ:
27.4, 45.6, 59.3, 128.5, 128.7, 128.7, 131.2, 131.2, 131.2. ESI-HRMS (m/z): calcd.
for C9H12NOPSNa [M+Na]+ : 236.0275; found 236.0270.
The preparation and workup of compounds 2-9 were similar to that of compound
1. The difference was using the corresponding amine instead of ethanolamine.
Compound 2:
Compound 2 was obtained as light yellow viscous oil. Yield: 42.4%. IR (KBr,
cm-1): 3265 (NH), 1436 (P=S), 1121 (C-O); 1H NMR (400 MHz, CDCl3) δ7.72 (s, 1H,
Ar-H), 7.45-7.28 (m, 4H, Ar-H), 3.92 (s, 1H, NH), 3.40 (q, J = 7.0 Hz, 1H, O-CH2),
3.16-3.05 (m, 1H, O-CH2), 2.08 (s, 1H, CH2), 1.14 (dt, J = 13.9, 7.9 Hz, 1H, CH2).13C
NMR (400 MHz, Chloroform-d) δ46.9, 67.6, 128.5, 128.7, 128.7, 131.2, 131.2, 131.2.
ESI-HRMS (m/z): calcd. for C8H10 NOPSNa [M+Na]+ : 222.0118; found 222.0121.
Compound 3:
The preparation and workup of compound 3 were similar to compound 1 except
1,3-propanediol (0.4 ml, 6.63×10-6 mol) were used instead of 3-aminopropanol.
Chromatography on silica gel with (PE/EA 5:1) as eluent gave the product. White oil
1
was obtained. Yield: 35.4%. IR (KBr, cm-1): 1440 (s,P=S), 1125 (s,C-O); H NMR
(400 MHz, Chloroform-d) δ7.88 (ddd, J = 14.6, 5.4, 3.4 Hz, 2H, Ar-H), 7.55 – 7.36
(m, 3H, Ar-H), 4.71 (dddd, J = 11.6, 9.9, 8.5, 3.0 Hz, 2H, O-CH2), 4.19 (ddt, J = 18.3,
11.5, 4.4 Hz, 2H, O-CH2), 2.18 (dtt, J = 14.3, 9.4, 4.3 Hz, 1H, CH2), 1.87 (dtq, J =
13
14.5, 4.9, 2.7 Hz, 1H, CH2); C NMR (400 MHz, Chloroform-d) δ26.2, 58.2, 58.2,
128.7, 128.7, 129.0, 129.0, 129.0, 131.2. ESI-HRMS (m/z): calcd. for C9H11O2PSNa
[M+Na]+ : 237.0115 ; found 237.0110.
Compound 4:
The preparation and workup of compound 4 were similar to compound 1 except
1,3-diaminopropane(0.4 ml, 6.63×10-6 mol) were used instead of 3-aminopropanol.
Chromatography on silica gel with (PE/EA 3:1) as eluent gave the product. A white
granular solid was obtained. Yield: 42.0%. IR (KBr,cm-1): 3252 (m, C-NH), 1436
(s,P=S); 1H NMR (400 MHz, Chloroform-d) δ7.80 (dtd, J = 13.5, 4.9, 2.7 Hz,2H, Ar-
H), 7.40 (dt, J = 5.3, 2.7 Hz, 3H, Ar-H), 3.31-3.15 (m, 2H, N-CH2), 3.05-2.91 (m, 2H,
N-CH2), 2.86 (s, 2H, NH), 1.82-1.66 (m, 1H, CH2), 1.46-1.35 (m, 1H, CH2); 13C
NMR (400 MHz, Chloroform-d) δ25.8, 58.2, 58.2, 128.7, 128.7, 129.0, 129.0, 129.0,
131.2. ESI-HRMS (m/z): calcd. for C9H14N2PS [M+H]+ : 212.0537 ; found 213.0567.
Compound 5:
The preparation and workup of compound 5 were similar to compound 1 except
ethylene glycol (0.4 ml, 6.63×10-6 mol) were used instead of 3-aminopropanol.
Chromatography on silica gel with (PE/EA 3:1) as eluent gave the product. A white

ACCEPTED MANUSCRIPT
granular solid was obtained. Yield: 56.4%. IR (KBr,cm-1): 1438.8 (s, P=S), 1118.2 (s,
1
C-O); H NMR (400 MHz, Chloroform-d) δ7.92-7.84 (m, 2H, Ar-H), 7.61-7.45 (m,
3H, Ar-H), 4.65-4.57 (m, 2H, O-CH2), 4.47-4.36 (m, 2H, O-CH2); 13C NMR (400
MHz, Chloroform-d) δ63.4, 63,4, 128.7, 128.7, 129.0, 129.0, 129.0, 131.2. ESI-
HRMS (m/z): calcd. for C8H9O2PSNa [M+Na]+:222.9959; found 222.9906.
Compound 6:
The preparation and workup of compound 6 were similar to compound 1 except
ethylenediamine (0.4 ml, 6.63×10-6 mol) were used instead of 3-aminopropanol.
Chromatography on silica gel with (PE/EA 3:1) as eluent gave the product. A yellow
granular solid was obtained. Yield: 35.0%. IR (KBr, cm-1): 3316.6 (m, C-NH), 1436.9
1
(s, P=S). H NMR (400 MHz, Chloroform-d) δ7.92 (ddd, J = 14.8, 8.2, 1.6 Hz, 2H,
Ar-H), 7.42-7.29 (m, 3H, Ar-H), 3.57-3.46 (m, 2H, CH2), 3.35-3.26 (m, 2H, CH2),
2.65 (d, J = 9.6 Hz, 2H, NH); 13C NMR (400 MHz, Chloroform-d) δ45.7, 45.7, 128.1,
128.1, 131.2, 131.2, 131.2, 132.0. ESI-HRMS (m/z): calcd. for C H11N2PSNa
8
[M+Na]+ : 221.0278; found 221.0266.
Compound 7:
The preparation and workup of compound 7 were similar to compound 1 except
2-(hydroxymethyl)pyrrolidine (0.6 ml, 6.63×10-6 mol) were used instead of 3-
aminopropanol. Chromatography on silica gel with (PE/EA 4:1) as eluent gave the
product. A white solid was obtained. Yield: 42.4%. IR (KBr, cm-1): 3447.8 (m, C-NH),
1
1436.9 (s, P=S), 1114.5 (s, C-O). H NMR (400 MHz, Chloroform-d) δ7.80 (dd, J =
15.0, 7.5 Hz, 2H, Ar-H), 7.41-7.25 (m, 3H, Ar-H), 4.23 (dddd, J = 22.3, 8.6, 6.2, 2.1
Hz, 1H, O-CH2), 4.00 (qd, J = 7.1, 3.9 Hz, 1H, O-CH2), 3.80-3.63 (m, 2H, N-CH2),
2.88-2.71 (m, 1H, N-CH), 1.98-1.87 (m, 1H, CH2), 1.83 (td, J = 8.2, 4.7 Hz, 2H, CH2),
1.62 (dt, J = 11.8, 5.3 Hz, 1H, CH2). 13C NMR (400 MHz, Chloroform-d) δ24.3, 28.8,
48.1, 64.0, 70.7, 128.5, 128.7, 128.7, 131.2, 131.2, 131.2.ESI-HRMS (m/z): calcd. for
C11H14NOPSNa [M+Na]+ : 262.0431; found 262.0369.
Compound 8:
The preparation and workup of compound 8 were similar to compound 1 except 4-
amino-1-butanol (0.4 ml, 6.63×10-6 mol) were used instead of 3-aminopropanol.
Chromatography on silica gel with (PE/EA 3:1) as eluent gave the product. A white
granular solid was obtained. Yield: 56.8%. IR (KBr, cm-1): 3418.0 (m, C-NH),
1
1436.9.8 (s, P=S), 1120.1 (s, C-O); H NMR (400 MHz, Chloroform-d) δ7.82 (ddt, J
= 14.0, 6.7, 1.6 Hz, 2H, Ar-H), 7.38 (dddt, J = 12.6, 8.5, 6.1, 1.9 Hz, 3H, Ar-H), 4.58-
4.46 (m, 1H, O-CH2), 4.21-4.05 (m, 1H, O-CH2), 3.13-2.96 (m, 1H, O-CH2), 2.80-
2.66 (m, 1H, O-CH2), 1.91-1.82 (m, 1H, NH), 1.78 (dp, J = 13.2, 3.5 Hz, 2H, CH2),
1.74-1.67 (m, 1H, CH2), 1.58-1.45 (m, 1H, CH2); 13C NMR (400 MHz, Chloroform-d)
δ27.8, 30.2, 37.5, 70.5, 128.5, 128.7, 128.7, 131.2, 131.2, 131.2. ESI-HRMS (m/z):
calcd. for C10H14NOPSNa [M+Na]+ : 250.0431; found 250.0429.