Convergent synthesis of novel muramyl dipeptide analogues: Inhibition of porphyromonas gingivalis-induced pro-inflammatory effects by high doses of muramyl dipeptide

Article abstract of DOI:10.1021/acs.jmedchem.6b00681

Porphyromonas gingivalis (P.g.)-induced TNF-α can be affected by muramyl dipeptide (MDP) in a biphasic concentration-dependent manner. We found that in P.g.-exposed macrophages, treatment with 10 μg/mL of MDP (MDP-low) up-regulated TNF-α by 29%, while 100

Full text of DOI:10.1021/acs.jmedchem.6b00681

Subscriber access provided by - Access paid by the | UCSB Libraries
Article
Convergent Synthesis of Novel Muramyl Dipeptide Analogs:
Inhibition of Porphyromonas gingivalis-induced Pro-
inflammatory Effects by High Doses of Muramyl Dipeptide
Bin Cai, James S. Panek, and Salomon Amar
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00681 • Publication Date (Web): 29 Jun 2016
Downloaded from http://pubs.acs.org on June 29, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Convergent Synthesis of Novel Muramyl Dipeptide
Analogs: Inhibition of Porphyromonas gingivalis-
induced Pro-inflammatory Effects by High Doses of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Muramyl Dipeptide
Bin Cai§, James S. Panek§* and Salomon Amar†*
†Center for AntiꢀInflammatory Therapeutics, Department of Molecular & Cell Biology, Boston
University Goldman School of Dental Medicine, 650 Albany Street, Boston, MA, 02118
§ Department of Chemistry, Boston University, Metcalf Center for Science and Engineering, 590
Commonwealth Avenue, Boston, MA, 02215
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 47
1
2
3
4
ABSTRACT
5
6
7
8
Porphyromonas gingivalis (P.g)ꢀinduced TNFꢀα can be affected by muramyl dipeptide (MDP) in
a biꢀphasic concentrationꢀdependent manner. We found that in P.gꢀexposed macrophages,
treatment with 10 µg/ml of MDP (MDPꢀlow) upꢀregulated TNFꢀα by 29%, while 100 µg/ml or
higher (MDPꢀhigh) significantly decreased it (16% to 38%). MDPꢀhigh was found to affect the
ubiquitinꢀediting enzyme A20 and activator protein 1 (AP1). An AP1 binding site was found in
the promoter region of A20. A20 promoter activity was upꢀregulated after transfection of AP1
cDNA in cells. Four analogs of MDP (3ꢀ6) were prepared through a convergent strategy
involving the synthesis of two unique carbohydrate fragments, 7a and 7b using the peptide
coupling reagents, EDCI and HOAt. Analog 4 improved MDP function and P.gꢀinduced
activities. We propose a new signaling pathway for TNFꢀα induction activated after exposing
macrophages to both P.g and MDPꢀhigh or analog 4.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 47
Journal of Medicinal Chemistry
1
2
3
4
1) INTRODUCTION
5
6
7
8
Two major classes of pattern recognition receptors (PRRs) include membraneꢀbound Tollꢀlike
receptors (TLRs) and soluble, cytosolic nucleotideꢀbinding oligomerization domain (NOD) like
receptors (NLRs)1. Where TLRs help to sense extracellular pathogens, NLRs provide
surveillance of the cytosol. Mammals have two closely related NOD family members, NOD1
and NOD2, both of which contain caspase recruitment domains (CARDs), a central NOD
(NACHT) domain, and Cꢀterminal Leuꢀrich repeats (LRRs)2ꢀ5. Studies have linked NOD2
mutations to severe inflammatory diseases6ꢀ7, suggesting that NOD2 might play a
pathophysiological role in inflammatory disorders. However it is important to note that chronic
stimulation of NOD2 was reported to mediate tolerance to bacterial products.6ꢀ7 The
mechanism(s) behind this tolerance remain poorly understood. Following breakdown of bacterial
cell walls, muramyl dipeptide (MDP) is formed and detected by NOD2. Upon detection of low
doses of MDP, NOD2 binds to the kinase receptorꢀinteracting protein 2 (RIP2) via CARDꢀ
CARD homophilic interactions—a necessary step for downstream signaling to proceed.8ꢀ10
Signaling to RIP2 leads to NFκB transcriptional activity through the inhibitor of κB (IκB) kinase
complex, as well as other cascades involving mitogenꢀactivated protein (MAP) kinases that
result in the production of proꢀinflammatory cytokines and chemokines such as interleukinꢀ6 (ILꢀ
6), TNFꢀα, ILꢀ12, and ILꢀ811. In contrast, a higher dose of MDP has been found to dampen the
inflammatory response12ꢀ13. Studies on NOD2 mutations suggest that NOD2 might mediate
tolerance to bacterial products, but the precise mechanisms have not yet been identified.6,14
The effects of MDP are biphasic: at 10 µg/ml (MDPꢀlow), MDP activates the inflammatory
process, while a dose of 100 µg/ml or higher (MDPꢀhigh) dampens the process by inhibiting the
NFkBꢀmediated cytokine response.7,15ꢀ16 No signs of cytotoxicity or apoptosis were observed in
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 47
1
2
vitro at high MDP doses15. While the proꢀinflammatory effects of low doses of MDP are well
documented, the mechanisms behind the antiꢀinflammatory effects of high doses of MDP remain
unclear. Understanding the dual actions of MDP on NOD2 may reveal an important signaling
pathway that may be therapeutically important in chronic inflammatory conditions.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The realization that MDP and its derivatives exhibit potent antiꢀinflammatory effects has
resulted in active research programs in both academia17 and industries. Mifamurtide, an analog
of MDP originally developed by Novartis and commercialized by Takeda, was approved in the
United States and Europe for the treatment of osteosarcoma. In reviewing studies aimed at
identifying proꢀinflammatory cytokine inhibitors, we learned that conformationally biased
(bicyclic systems) MDP derivatives serve as new lead compounds accessed through an efficient
convergent synthesis. A literature search revealed that existing methods did not describe the
preparation of conformationally biased, bicyclic analogs18ꢀ21 which are the first examples of this
therapeutic class. In addition, few reports19ꢀ20,22ꢀ25 have fully characterized synthetic
intermediates and final products, while also allowing for their preparation in quantities that
would be sufficient for biological evaluation18ꢀ20,26 (see Supporting Information for details). This
work enables the production of stereochemically pure compounds in useful quantities: >20 grams
of the Nꢀacetyl sugar fragment in a fourꢀstep sequence, requiring a single purification step and
features an efficient, sevenꢀstep route to the associated dipeptides.
With this chemical strategy in place, we looked for the signal transduction pathway associated
with the biochemical effects of MDPꢀhigh on inflammation. Based on previous studies showing
that the ubiquitinꢀediting enzyme A20 functions as an antiꢀinflammatory protein27 and as a key
negative regulator of NFκB, 28 we proposed a role of A20 in MDP antiꢀinflammatory function.
ACS Paragon Plus Environment

Page 5 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Four enantiomerically enriched analogs of MDP were synthesized and among them, analog 4
exhibited the most potent activity. A new signaling pathway mediated by MDPꢀhigh or analog 4
after Porphyromonas gingivalis (P.g) stimulation is hereto advocated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2) RESULTS AND DISCUSSION
2.1) MDP effects on P.gꢀinduced TNFꢀα, A20, and other factors
Our previous data indicated that LPSꢀinduced TNFꢀα in macrophages was increased by
concurrent treatment with MDPꢀlow treatment but surprisingly, was decreased by MDPꢀhigh
treatment7 29. To further confirm this, a dose course of MDP was performed. As shown in Figure
1, MDPꢀlow upꢀregulated P.gꢀinduced TNFꢀα by 29% but MDPꢀhigh significantly decreased
P.gꢀinduced TNFꢀα secretion by 16% to 38% in macrophages. Furthermore, in order to see
whether a higher dose of MDP had an immediate toxic effect on cell survival, a toxicity test was
performed. As shown in Figure 2, neither the low nor highꢀdose MDP treatments significantly
affected macrophage survival, even when a much higher dose (1,000 µg/ml) of MDP was used.
Meanwhile, we also found that a high dose of MDP in macrophages did not affect NOD2 or
RIP2 transcription (Figure 3AꢀD). However, this treatment significantly induced A20 with a 5.6ꢀ
fold increase in mRNA levels, as compared to the control, while MDPꢀlow treatment did not
(Figure 3E). This phenomenon was further confirmed by Western blot analysis (Figure 4).
Interestingly, we also found that this treatment upꢀregulated the activator protein 1 (AP1)
production but not MyD88, NFκB (Figure 4A), p53, or p21 (Figure 4B), compared to the control.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. ELISA analysis of TNFꢀα expression. The preꢀcultured primary macrophages from wildꢀtype (WT) mice were untreated or treated with
10, 100, 1,000 µg/ml of MDP as the control, or treated with P.g alone as a positive control, or coꢀtreated with P.g plus 10, 100, or 1,000 µg/ml of
MDP for 3 h. The media from each test group were used to detect TNFꢀα via an ELISA kit (Invitrogen). Data are represented as Mean ±S.E.
*p≤0.05.
Figure 2. Toxicity test. WT mouse macrophages were treated with different quantities of MDP for 0, 2, 4, 8, 12, 16, 18, 24 h. Cells from each
time point were collected by Trypsin. Cells were stained by Trypan blue and their survival rate was calculated. The survival rate by 0 µg/ml of
MDP for 24 h was assigned a value of 100% as the baseline, to which the actual values of the others was compared. Data are represented as Mean
±S.E. *p≤0.05.
Figure 3. RTꢀPCR analysis of MDPꢀmediated gene expression. WT mouse macrophages were untreated as the negative control, P.g alone as the
positive control, or coꢀtreated with P.g plus MDPꢀlow treatment, or MDPꢀhigh treatment for 3 h. Total mRNA from each test group were
assessed by RTꢀPCR with the primers of NOD2, RIP2, or A20 and normalized with GAPDH. The intensity of mRNA levels of NOD2, or RIP2
from the P.g alone treatment (AꢀD) were assigned to a base value (100%). For A20, the intensity of mRNA levels after MDPꢀlow treatment (E)
was assigned to a base value (100%). The others were calculated relative to this base value. Triplicate assays were conducted. Mean SEM.
ACS Paragon Plus Environment

Page 7 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Western blot analysis. WT mouse macrophages were untreated or treated with a SC peptide as the negative control, MDPꢀlow
treatment, or MDPꢀhigh treatment for 3 h. The total protein from each test group was assessed by Western blot with antibodies (Abs) acting
against AP1, A20, MyD88, NFkB, or Actin as the control. The intensity of protein levels by treatment of a SC peptide was assigned a value of
100% as the baseline, to which the actual value of others was compared (A). WT mouse macrophages were treated with either different doses of
MDP (0, 10, 100, 1,000 µg/ml), or coꢀtreated with P.g plus different doses of MDP (0, 10, 100, 1,000 µg/ml) for 3 h. The total protein from each
test group was assessed by Western blot with Abs against p21, p53, or Actin as the control (B). Triplicate assays are presented.
2.2) Activation of A20 via AP1 protein/DNA interaction
Because MDPꢀhigh treatment in cells induced both A20 and AP1, we hypothesize that they are
associated in the response to MDP. Through further analysis of the A20 DNA sequence, we
found that the promoter region of A20 contains an AP1 binding site (Figure 5A). In order to
determine whether AP1 interacts with A20, a ChIP analysis assay using RAW 264.7 cells was
performed. As shown in Figure 5B, a PCR amplification of the A20 promoter DNA was
observed when cells were treated with MDPꢀhigh but not MDPꢀlow, compared to the control
(transfected with AP1 cDNA). To determine the region of the A20 promoter important for AP1
binding activity, RAW 264.7 cells were coꢀtransfected with A20 promoter DNA and different
concentrations of AP1 cDNA, after which the protein from each test group was assessed via a
luciferase assay. As shown in Figure 6, A20 promoter activity increased doseꢀdependently with
increasing AP1 cDNA concentration, indicating that AP1 is an important factor for upꢀregulating
A20 via protein/DNA interactions. As it is well known that activation of A20 restricts NOD2,30ꢀ
31
we therefore propose a new signaling pathway mediated by high doses of MDP treatment in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 47
1
2
3
4
5
6
7
8
9
response to P.g (Figure 7). Specifically, we suggest that through this pathway, when cells are
treated with a high dose of MDP, MDP is transferred into the cytoplasm and activates Jun Nꢀ
terminal kinases (JNKs);7 JNKs then upꢀregulate AP1, which in turn activates A20 and restricts
NOD2 ubiquitination, consequently inhibiting TNFꢀα secretion that would otherwise have been
induced in response to P.g. We speculate that the signal transduction pathway identified for
MDPꢀhigh would be the same as the one activated by analog 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Chromatin Immunoprecipitation (ChIP) assay. Diagram of the A20 promoter region sequences, with the location of the AP1 binding
site and PCR primers of A20 (A). DNA fragments of A20 or GAPDH as controls were amplified by PCR, with Input or IPs after the transfection
of a full length AP1 cDNA in RAW264.7 cells (B). The assays were performed in triplicate and a representative experiment is presented.
Figure 6. Promoter assay. RAW264.7 cells were transfected with pcDN3 (lane 1), 0.2 µg of the A20 promoter DNA alone (lane 2) as the control,
coꢀtransfected with the A20 promoter DNA (0.2 µg) plus different quantities of the full length AP1 cDNA (0.1 µg in lane 3, 0.25 µg in lane 4, 0.5
µg in lane 5). Proteins extracted from each group were measured by luciferase assay (A) and by Western blot (B), with antibodies acting against
AP1 or Actin as the control. Triplicate assays are presented.
ACS Paragon Plus Environment

Page 9 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. MDPꢀhighꢀinduced signaling pathway in response to P.g. MDP will be transferred into the cell’s’ cytoplasm and will activate JNKs,
which will upꢀregulate AP1. AP1 will activate A20 and restrict NOD2 ubiquitin, consequently inhibiting TNFꢀα secretion in response to P.g.
2.3) Synthesis of novel MDP analogs
Since the discovery of the key role of MDP in NOD2ꢀmediated innate immune responses to
bacterial invasions in certain inflammatory diseases, research on the mechanism of actions and
signaling pathways have been flourishing.17,32ꢀ37 More recently, we have shown that treatment
with higher concentration of MDP could have antiꢀinflammatory effects through regulating
NOD2, mitigating atherosclerosis and bone loss. In a related study, the Kobayashi group38
studied four MDP analogs and identified one aminosaccharide compound (DFK1012) which
could inhibit the production of proꢀinflammatory cytokines upon stimulation of the immune
receptor proteins, TLR and/or NLR. DFK1012 has also proved to have no cytotoxicity, thus
exhibiting promising antiꢀinflammatory effects. Since then, there have surprisingly been no
further reports on the SAR of MDP pertaining to its antiꢀinflammatory effects. This
underdeveloped, yet promising field prompted us to study the SAR of selected MDP analogs by
employing a convergent synthesis strategy delineating the signaling pathway to analogs, and the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 47
1
2
3
4
5
6
eventual identification of a lead compound. In that context, we targeted four novel analogs
(Figure 8) because of their synthetic accessibility.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. MDP and synthetic analogs of MDP
The underlying principles that guided us to design an early stage probe into the SAR were that
the analogs (1) needed to be synthetically accessible, (2) must bear the features of both MDP and
DFK1012, (3) and yet be differentiated from each other in subtle substitutions in the
carbohydrate fragments. As such, we selected to independently modify the dipeptide and
carbohydrate fragments before merging the two. In that regard, two variables were considered:
the stereocenter on the lactic acid moiety and the rigidity the Nꢀacetylmuramic acid fragment by
incorporation of the cyclic ketal on the C4 and C6 hydroxyl groups, with the expectation that a
conformationally biased carbohydrate fragment may provide guidance for further structural
ACS Paragon Plus Environment

Page 11 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
optimization. As in Figure 8, analog 3 is devoid of the methylꢀbearing stereocenter on the lactic
acid subunit, while analog 6 contains the natural stereochemistry of the lactic acid moiety.
Similarly, analog 5 was produced with the cyclic ketal and lastly, analog 4 merged the two
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
variables—the (4R)ꢀstereocenter of the lactic acid group and the bicyclic ketal.
Scheme 1. Retrosynthetic analysis of MDP analogs
In our initial efforts toward the synthesis of MDP analogs, progress was hampered by low
yielding reactions.23ꢀ24,39ꢀ40 More importantly, these methods didn’t allow for the preparation of
the dipeptide fragment from simple, inexpensive, and/or commercially available materials. In
that context, the Grimes group26 described a synthetic approach leading to MDPꢀlike agents that
cleverly employed the functional group interconversion strategy of an azide to Nꢀacylated
variants at the C2ꢀposition, and subsequently studied their effects on NOD2 signaling and
stability. The limitations of their synthesis include the production of mixtures of α/βꢀanomers of
the carbohydrate fragments, and the inefficient route to the dipeptide.
Our approach focused on establishing a reliable, scalable, and divergent solutionꢀphase
synthesis of diastereomerically and enantiomerically pure MDP analogs in order to allow for the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 47
1
2
3
4
5
6
7
8
9
preparation of MDP analogs in useful quantities for evaluating P.gꢀinduced proꢀinflammatory
cytokine inhibitors. The present work marks the first reliable preparation of this dipeptide
structural type with complete spectroscopic characterization of all intermediates and final
products (see Supporting Information for details). In addition, the synthesis of carbohydrate
fragments produced Nꢀacetyl glucosamine scaffolds as single αꢀanomers.22,41 Retrosynthetically,
analogs 6 and 3 could be derived respectively through removal of the ketal of 4 and 5 (Scheme
1). Analogs 4 and 5 could be constructed by standard peptide coupling between carboxylic acids
7a and 7b and dipeptide 8, which were synthesized respectively from commercially available Nꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
acetylꢀDꢀglucosamine and Dꢀglutamic acid.
Scheme 2. Synthesis of carboxylic acids 7a and 7ba
The synthesis of the carboxylic acids 7a and 7b commenced with the introduction of the
benzyl ether at the anomeric position of Nꢀacetylꢀ ꢀglucosamine (Scheme 2). Originally, we
D
adopted a twoꢀstep sequence similar to that reported by Hecker et al.,22 wherein the Nꢀacetyl
amino sugar 10 was produced by using intermediate 9 without purification. However, in our
hands, we opted to purify 9 by recrystallizing it in hot ethanol before moving on to the next steps
ACS Paragon Plus Environment

Page 13 of 47
Journal of Medicinal Chemistry
1
2
and obtained 9 as a single αꢀanomer.42 Thus, 9 was produced by heating Nꢀacetylꢀ
Dꢀglucosamine
3
4
5
6
7
8
in benzyl alcohol at 95oC, in the presence of anhydrous hydrochloric acid (1.0 equiv, 4N solution
in 1,4ꢀdioxane) in tetrahydrofuran (THF) as the promoter.22 The orientation of the benzyl ether
was assigned as axial based on the coupling constant J=3.5 Hz at δ=4.68 (characteristic of a cis
axialꢀequatorial orientation) for the anomeric proton. The stereochemical assignment was further
confirmed by Heteronuclear Single Quantum Coherence (HSQC) and Heteronuclear Multiple
Bond Correlation (HMBC) nuclear magnetic resonance spectroscopic experiments (see
Supporting Information for details). Protection of 9 with 2,2ꢀdimethoxypropane and a catalytic
amount of pꢀtoluenesulfonic acid monohydrate (pꢀTsOH) in acetone afforded acetonide 10 with
100% conversion,22 which was then used without further purification. Oꢀalkylation of 10 with
(S)ꢀ2ꢀchloropropanoic acid22 and bromoacetic acid using sodium hydride (3 equiv and 4 equiv,
respectively) as a base in THF cleanly produced alkylation products 7a and 7b respectively. The
production of 7a as a single diastereomer was confirmed by comparing optical rotation of 7a to
the literature value ([α]ꢀꢁꢂ = +118 (c 1.0, CHCl3), literature [α]ꢀꢁꢃ = +117 (c 1.0, CHCl3).22
Notably, this threeꢀstep sequence required only one single purification step employing a
conventional chromatography on SiO2 (silica gel). Accordingly, we established a facile and
scalable route that allowed for the preparation of the functionalized glucosamines 7a and 7b in
multiꢀgram quantities (>20 grams) with an overall yield of 43% (see Supporting Information for
details).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The preparation of dipeptide 17 was guided by a desire to establish a reproducible route
suitable for multiꢀgram production of the diastereomerically and enantiomerically pure material
(Scheme 3). A disconnection at the peptide bond revealed that the dipeptide could be constructed
by peptide coupling between the trifluoroacetic acid (TFA) salt of isoglutamine benzyl ester 16
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 47
1
2
3
4
and Bocꢀ
Lꢀalanine. Despite the fact that amino acid 16 is a substructure of the proposed MDP
5
6
7
8
analogs, existing routes do not permit a gramꢀscale synthesis, and thus present an opportunity for
further development.23,25ꢀ26 In that regard, our synthesis commenced with a modified threeꢀstep
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
sequence originally reported by Ressler.25 Commercially available
Dꢀglutamic acid was protected
as benzyl carbamate 11, according to a published report by Abell43, wherein the pH was carefully
maintained in the range of 8 to 10 by using a combination of sodium carbonate and sodium
bicarbonate in a solution of acetone and water (1:10). This crucial pH range minimized the
decomposition of benzylchloroformate in an aqueous solution and prevented racemization of the
derived carbamate. In our hands, the largest scale of this reaction in one pot was 10 grams with
an excellent yield of 92%. It was postulated that the αꢀcarboxylic acid of the bisꢀcarboxylic acid
11 would exhibit a greater acidity than the γꢀcarboxylic acid, as evidenced by a smaller chemical
shift of the αꢀcarbon in the 13CꢀNMR. Based on this assumption, a regioselective amidation of 11
was achieved by selectively converting the αꢀcarboxylic acid into a mixed anhydride using
isobutylchloroformate and triethylamine.25 Isoglutamine 13 was obtained as a hydrochloric acid
salt by hydrogenolysis of the carboxybenzyl (Cbz) protecting group of 12 and subsequent
recrystallization with concentrated hydrochloric acid. Isoglutamine benzyl ester 15 was
generated in good yield through benzylation of amide 14, which was produced by protection of
the isoglutamine hydrochloric acid salt 13 with Boc anhydride.44 This optimized route enabled
the preparation of multiꢀgram quantities (>20 grams) of 15 in four steps, requiring only one silica
gel chromatography operation. Subsequent removal of the tertꢀbutyloxycarbonyl (Boc)
44
protecting group of 15 using TFA (13 equiv.) afforded 16, setting the stage for the peptide
coupling between 16 and BocꢀL
ꢀalanine. Optimization of coupling conditions45ꢀ48 identified the
combination of 1ꢀethylꢀ3ꢀ(3ꢀdimethylaminopropyl)carbodiimide (EDCI), 1ꢀhydroxyꢀ7ꢀ
ACS Paragon Plus Environment

Page 15 of 47
Journal of Medicinal Chemistry
1
2
3
4
azabenzotriazole (HOAt), and 2,2,6,6ꢀtetramethylpiperidine (TMP) in dichloromethane (CH2Cl2)
5
6
as optimal. Final deprotection of 17 afforded the primary amine used for the coupling of 8 and
7
8
9
protected Nꢀacetyl glucosamine 7a and 7b.
Scheme 3. Synthesis of dipeptide 8a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
With 8, 7a, and 7b in hand, our attention turned to the fragment coupling between 8 and
carbohydrate 7a and 7b (Scheme 4). Gratifyingly, by employing the conditions of EDCI, HOAt,
and TMP in CH2Cl2, 4 and 5 were formed cleanly after twelve hours. Analogs 6 and 3 were
obtained by treating 4 and 5 with TFA in dichloromethane for one hour. 49
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 47
1
2
3
Scheme 4. Fragment couplings to access MDP analogsa
4
5
6
7
Me
HO
HO
O
O
Me
O
O
O
O
O
O
Me
NH
Me
NH
NH
O
Me
OBn
a)
b)
NH
O
OBn
Me
O
Me
8
O
Me
H2N
NH
H2N
NH
9
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OBn
OBn
O
O
4
6
8
Me
HO
HO
O
O
O
Me
O
O
O
O
O
Me
NH
Me
NH
NH
O
OBn
c)
d)
NH
O
OBn
O
H2N
Me
O
H2N
Me
NH
NH
O
O
OBn
OBn
O
O
5
3
aReagents and conditions: (a) 7a, TMP ,EDCI, HOAt, CH2Cl2, r.t. 12h, 56%, two steps; (b) TFA, CH2Cl2, 0
oC, 1h, 73%; (c) 7b, TMP ,EDCI, HOAt, CH2Cl2, r.t. 12h, 57%, two steps; (d) TFA, CH2Cl2, 0 oC, 1h, 63%.
2.5) Analysis of MDP analogs
MDP was shown earlier to express activity inhibiting the synthesis of TNFꢀα7. Therefore, we
were interested in MDPꢀlike compounds with improved antiꢀinflammatory potency (more
typically nonspecific ester hydrolysis). Four MDP derivatives (3ꢀ6) were synthesized (Figure 8B)
and further analyzed (Figures 9ꢀ11). As shown in Figures 9 and 10, among the four derivatives,
only 4 was similar to MDP, in that it gradually reduced TNFꢀα as a result of an increased
concentration and induced AP1/A20 gene expression. We also found that a high dose of analog 4
in macrophages reduced the transcription of RIP2, NOD2, and NFκB to an even greater extent
than MDP treatment did (Figure 11). This indicates that, based on its efficacy, analog 4 is indeed
an improved compound in comparison to MDP. To confirm our in vitro observations, in vivo
P.gꢀinduced endotoxic shock was performed50. As shown in Figure 12, P.g-injectionꢀonly
induced septic shock, while death was precipitated with additional treatment of MDPꢀlow.
ACS Paragon Plus Environment

Page 17 of 47
Journal of Medicinal Chemistry
1
2
3
4
However, injecting mice with MDPꢀhigh or analog 4 significantly reduced endotoxic shock and
5
6
further prevented death.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. ELISA analysis of individual MDP or its derivativeꢀmediated TNFꢀα gene expression. WT mouse macrophages were untreated or
treated with P.g alone or the different MDP derivatives alone as the controls or coꢀtreated with P.g plus 10, 100, 200 µg/ml of MDP or analogs 3ꢀ
6 for 3 h. The media from each test group were used to detect TNFꢀα via an ELISA kit. Data are represented as Mean ±S.E. *≤0.05.
Figure 10. ELISA analysis of TNFꢀα production in response to P.g and/or MDP, analog 3, or analog 4. WT mouse macrophages were untreated
or treated with P.g, MDP, analog 3, or analog 4 as controls, or coꢀtreated with P.g plus 10, 100, 200 µg/ml of MDP, analog 3, or analog 4 for 3 h.
The supernatants from each group were assessed by ELISA to detect the levels of TNFꢀα protein (A). The regulatory proteins from each group
were detected by Western blot (B) using antibodies against AP1, A20, NFkB, or Actin as the control. Data were represented as Mean ± S.E.
*p<0.05.
Figure 11. RTꢀPCR analysis of individual MDP or derivativeꢀmediated gene expression. WT mouse macrophages were untreated or treated with
P.g alone, or MDP, analog 3, or analog 4 alone as the controls or coꢀtreated with P.g plus 10, 100, 200 µg/ml of MDP, analog 3, or analog 4 for 3
h. The total RNAs from each group were assessed by RTꢀPCR with the primers of NFkB (A), RIP2 (B), or NOD2 (C). GAPDH was used as a
house keeping control gene for normalization. Data are represented as Mean ± S.E. *p<0.05.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 12. Analysis of endotoxic shock. WT mice were i.p. injected with 2 mg/kg (50 µg) of MDPꢀlow (A) or the derivative analog 4 (B); 8
mg/kg (200 µg) of MDPꢀhigh (A) or analog 4 (B); or DMSO alone as the negative control. Following injection, they were immediately i.v.
injected with P.g (5x108 cells/mouse). Mice were then monitored and their status was recorded. The pꢀvalue for the test groups (n=3) was
calculated by the Wald’s test in a discrete time hazard model using version 9.2 of SAS software gene for normalization. Data are represented as
Mean ± S.E. *p<0.05.
3) CONCLUSIONS
Inflammatory and immune responses appear to be important mediators linking infection and
systemic conditions, such as atherogenesis, possibly requiring residence of the pathogen in the
vessel wall. Interestingly, the majority of pathogens associated with atherogenesis share four
important characteristics: they are involved in common infections51; they are all intracellular
pathogens52; they establish longꢀterm persistent infections53; and they induce longꢀlasting
increases in specific antibody production54. Specific cytosolic recognition molecules (e.g.
NOD2) activated by MDP appear to be involved in this process, such that the infection itself is
appropriate to drive the activation of signaling pathways.55 We evaluated the role of NOD2 in
two chronic inflammatory diseases: atherosclerosis and alveolar bone loss. P.gꢀchallenged
ApoEꢀ/ꢀ mice injected with high doses of MDP displayed a reduction of serum cholesterol and
inflammatory cytokines, alveolar bone loss, and atherosclerotic lesions, compared to the P.g
challenged ApoEꢀ/ꢀ mice injected with saline7. In the present study we present evidence
ACS Paragon Plus Environment

Page 19 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
highlighting a high dose MDPꢀdependent signaling pathway which activates JNKs, induces AP1,
upꢀregulates A20 expression, restricts NOD2, inhibits NFκB, and consequently, reduces P.gꢀ
induced TNFꢀα production in mouse macrophages. Furthermore, using the power of modern
synthetic chemistry, we have described an efficient, highly convergent synthesis of
enantiomerically enriched MDP analogs (3ꢀ6) by merging the suitably functionalized
carbohydrate portion with the dipeptide fragment. The conformationally biased analog 4 was
found to further improve MDP function and inhibition of P.gꢀinduced proꢀinflammatory
activities. A novel signaling pathway leading to TNFꢀα expression that is influenced by exposing
macrophages to both P.g and a high dose of MDP or analog 4 is hereto advocated. MDPꢀhigh
and its optimized analog 4 both reduced P.gꢀinduced septic shock and death—a clear indication
of their potent antiꢀinflammatory effect. We propose that the rigid bicyclic ring system present
on analog 4 is responsible for the enhanced reduction of TNFꢀα production and immunity to
septic shock and death. The present work paves the way for further MDP optimization to
improve its efficacy and safety.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4) EXPERIMENTAL SECTION
(A) Chemistry.
All reactions were carried out in oven or flameꢀdried glassware under argon atmosphere unless
otherwise specified. Triethylamine and 2,2,6,6ꢀtetramethylpiperidine were distilled over calcium
hydride and stored over potassium hydroxide. Dichloromethane and tetrahydrofuran were
obtained from a dry solvent system (alumina) and used without further drying. All other reagents
were used as supplied. Unless otherwise noted, reactions were magnetically stirred and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 47
1
2
3
4
5
6
7
8
9
monitored by thin layer chromatography with Macherey Nagel Polygram 0.20 mm silica gel 60
Å plates. Flash chromatography was performed on Sorbent Technologies 32ꢀ63 ꢁm 60 Å silica
gel. Yields refer to chromatographically and spectroscopically pure compounds, unless otherwise
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
13
noted. H and C NMR spectra were taken in CDCl3, DMSOꢀd6, and MeOHꢀd4 at 400 or 500
MHz (as indicated), respectively. Chemical shifts are reported in parts per million using the
solvent internal standard (chloroform, 7.24 and 77.0 ppm, DMSO, 2.5 and 40.0 ppm, and MeOH,
3.31 and 49.0 ppm respectively). Data are reported as follows: chemical shift, multiplicity (s =
singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad), coupling constant,
integration. Infrared resonance spectra were recorded on a Nexus 670 FTꢀIR spectrometer.
Optical rotations were recorded on a Rudolph Autopol II digital polarimeter at 589 nm and
reported as follows: [α]ꢁꢀꢄ (concentration in g/100 mL solvent and solvent). High resolution
massꢀspectra were obtained on a Waters QꢀTOF Mass Spectrometer at the Boston University
Chemical Instrumentation Center. The purity of all compounds were determined to be >95% by
UPLCꢀMS.
Nꢀ((2S,3R,4R,5S,6R)ꢀ2ꢀ(benzyloxy)ꢀ4,5ꢀdihydroxyꢀ6ꢀ(hydroxymethyl)tetrahydroꢀ2Hꢀ
pyranꢀ3ꢀyl)acetamide (9): NꢀacetylꢀDꢀglucosamine (20 g, 9.04 mmol, 1.0 equiv) was suspended
in 152 mL benzyl alcohol under argon at room temperature. Hydrogen chloride solution (4M in
anhydrous 1,4ꢀdioxane, 2.5 mL, 9.94 mmol, 1.1 equiv) was added dropꢀwise via a syringe. The
o
mixture was heated at 95 C for 4 h. After the reaction was complete, it was cooled down to
room temperature, and then poured into 1L cold ether with vigorous stirring. The resulting brown
cake was filtered, collected, and washed with cold ether. Purification by recrystallization from
ethanol afforded 9 as a white solid as a single αꢀanomer (18g, 64% yield). [α]ꢁꢀꢄ= +153.2 (c 1.0,
DMSO);
ACS Paragon Plus Environment

Page 21 of 47
Journal of Medicinal Chemistry
1
2
1H NMR (500 MHz, DMSOꢀd6): δ7.80 (d, J = 8.5Hz, 1H), 7.36ꢀ7.25 (m, 5H), 5.00 (d, J =
6Hz, 1H), 4.71 (d, J = 6Hz, 1H), 4.68 (d, J = 3.5Hz, 1H, anomeric), 4.65 (d, J = 12.5Hz, 1H),
4.52 (t, J = 5.7Hz, 1H), 4.40 (d, J = 12.5 Hz, 1H), 3.68ꢀ3.62 (m, 2H), 3.53ꢀ3.42 (m, 3H), 3.16ꢀ
3.11 (m, 1H), 1.81 (s, 3H);
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13C NMR (125 MHz, DMSOꢀd6): δ169.90, 138.37, 128.62, 127.95, 127.86, 96.33, 73.56,
71.34, 71.02, 68.13, 61.30, 54.20, 23.00;
IR (neat/cmꢀ1) υmax : 3288, 2908, 1637, 1549, 1124, 1039
HRMS (ESI) m/z calcd for C8H15NO6 [M+H] + 312.1447, found 312.1443;
UPLC TR 0.72 min.
Nꢀ((4aR,6S,7R,8R,8aS)ꢀ6ꢀ(benzyloxy)ꢀ8ꢀhydroxyꢀ2,2ꢀdimethylhexahydropyrano[3,2ꢀ
d][1,3]dioxinꢀ7ꢀyl)acetamide (10): 9(18 g, 57.8 mmol, 1 equiv) was suspended in 288 mL
acetone under argon. 2,2ꢀdimethoxypropane (72 mL,578.2 mmol, 10 equiv) was added in,
followed by pꢀtoluenesulfonic acid monohydrate (pꢀTSA·H2O) (1.09g, 5.76 mmol, 0.1 equiv).
The mixture was stirred at room temperature for 12 h. The reaction was quenched and
neutralized with saturated sodium bicarbonate solution. The aqueous layer was extracted with
dichloromethane (3 x 500 mL), and the combined organic layers were dried over MgSO4,
filtered, and concentrated to obtain pure 10 (20g, 100% yield). [α]ꢁꢀꢄ= +86.4 (c 1.0, CH2Cl2);
1H NMR (500 MHz, CDCl3): δ7.40ꢀ7.30 (m, 5H), 5.86 (d, J = 8.5Hz, 1H), 4.88 (d, J = 3.5Hz,
1H, anomeric), 4.71 (d, J = 12.0Hz, 1H), 4.45 (d, J = 12.0Hz, 1H), 4.17 (dt, J = 10.0, 4.0Hz,
1H), 3.84ꢀ3.60 (m, 5H), 1.98 (s, 3H), 1.52 (s, 3H), 1.43 (s, 3H);
13C NMR (125 MHz, CDCl3): δ171.41, 136.79, 128.65, 128.27, 128.07, 99.87, 97.16, 74.70,
70.95, 69.83, 63.66, 62.20, 54.08, 29.07, 23.25, 19.08;
IR (neat/cmꢀ1) υmax : 3319, 1656, 1546, 1372, 1267, 1199, 1122, 1073;
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 47
1
2
HRMS (ESI) m/z calcd for C18H25NO6 [M+Na] + 374.1580, found 374.1589;
UPLC TR 1.09 min.
3
4
5
6
7
8
9
(R)ꢀ2ꢀ(((4aR,6S,7R,8R,8aS)ꢀ7ꢀacetamidoꢀ6ꢀ(benzyloxy)ꢀ2,2ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dimethylhexahydropyrano[3,2ꢀd][1,3]dioxinꢀ8ꢀyl)oxy)propanoic acid (7a): Sodium hydride
(1.5g, 60% dispersion in mineral oil, 37.5 mmol, 3 equiv) was washed with hexanes (3 mL) three
time to remove mineral oil, and was suspended in 42 mL THF under argon. 10 (4.4g, 12.5 mmol,
1 equiv) was added into the above solution slowly, followed by (S)ꢀ2ꢀchloropropanoic acid (1.3
mL, 15.0 mmol, 1.2 equiv) dropꢀwise. The resulting mixture was stirred vigorously at room
temperature until there was no gas evolution. Then it was heated at 50 oC for 18 h (max speed of
stirring). The reaction mixture was cooled down to room temperature, quenched with 40mL DI
water, and acidified with phosphoric acid to pH=3. It was then extracted with dichloromethane (3
x 300 mL), and the combined organic layers were dried over MgSO4, filtered, and concentrated.
Purification over silica gel chromatography (50% EtOAc/CH2Cl2, then 4% MeOH/CH2Cl2)
afforded product 7a as a fluffy, white solid (4.5g, 85% yield). [α]ꢁꢀꢄ= +118.4 (c 1.0, CH2Cl2);
1H NMR (500 MHz, CDCl3): δ7.45 (d, J = 5.2Hz, 1H), 7.35ꢀ7.28 (m, 5H), 5.26 (d, J = 3.6Hz,
1H, anomeric), 4.65 (d, J = 11.9Hz, 1H), 4.47 (d, J = 11.9Hz, 1H), 4.40 (q, J = 7.1Hz, 1H),
3.94ꢀ3.90 (m, 1H), 3.82ꢀ3.65 (m, 5H), 2.00 (s, 3H), 1.51 (s, 3H), 1.42 (d, J = 7.1Hz, 3H), 1.40 (s,
3H);
13C NMR (125 MHz, CDCl3): δ177.07, 172.49, 137.21, 128.39, 127.89, 127.87, 99.57, 96.78,
75.77, 75.38, 75.24, 70.11, 63.77, 62.43, 54.56, 29.16, 22.60, 19.24, 18.73;
IR (neat/cmꢀ1) υmax : 3298, 2992, 2942, 2880, 1716, 1621, 1565, 1455, 1383, 1267, 1237, 1202,
1175, 1149, 1123, 1076, 1046;
HRMS (ESI) m/z calcd for C21H29NO8 [M+Na] + 446.1791, found 446.1794;
ACS Paragon Plus Environment

Page 23 of 47
Journal of Medicinal Chemistry
1
2
3
4
UPLC TR 1.30 min.
5
6
7
8
2ꢀ(((4aR,6S,7R,8R,8aS)ꢀ7ꢀacetamidoꢀ6ꢀ(benzyloxy)ꢀ2,2ꢀdimethylhexahydropyrano[3,2ꢀ
d][1,3]dioxinꢀ8ꢀyl)oxy)acetic acid (7b): Sodium hydride (2.19g, 60% dispersion in mineral oil,
45.6 mmol, 4 equiv) was washed with hexanes (4 mL) three time to remove mineral oil, and was
suspended in 60 mL THF under argon. 10 (4.0 g, 11.4 mmol, 1 equiv) was added into the above
solution slowly, followed by bromoacetic acid (2.225 g, 16.0 mmol, 1.4 equiv) dropwise. The
resulting mixture was stirred vigorously at room temperature until no gas evolution. Then it was
heated at 50oC for 18 h (max speed of stirring). The reaction mixture was cooled down to room
temperature, quenched with 100mL DI water, and acidified with phosphoric acid to pH=3. It was
then extracted with dichloromethane (3 x 300 mL), and the combined organic layers were dried
over MgSO4, filtered, and concentrated. Purification over silica gel chromatography (3.7%
MeOH/CH2Cl2) afforded product 7b as a fluffy, white solid (3.9 g, 84% yield). [α]ꢁꢀꢄ= + 93.2 (c
1.0, CH2Cl2);
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (500 MHz, CDCl3): δ7.36ꢀ7.28 (m, 5H), 7.24 (d, J = 6.3Hz, 1H), 5.16 (d, J = 3.8 Hz,
1H, anomeric), 4.67 (d, J = 11.9Hz, 1H), 4.47 (d, J = 11.9 Hz, 1H), 4.31 (ABq, J = 17.6Hz, 2H),
4.01 (ddd, J = 10.2, 6.6, 3.8 Hz, 1H), 3.82ꢀ3.65 (m, 5H), 2.00 (s, 3H), 1.51 (s, 3H), 1.41 (s, 3H);
13C NMR (125 MHz, CDCl3): δ174.26, 172.53, 137.18, 128.42, 127.91, 127.84, 99.68, 96.74,
76.06, 75.58, 70.04, 68.32, 63.72, 62.30, 54.39, 29.11, 22.57, 19.18;
IR (neat/cmꢀ1) υmax : 3309, 2995, 2932, 2882, 1719, 1621, 1560, 1433, 1383, 1267, 1233, 1201,
1174, 1146, 1119, 1074, 1042;
HRMS (ESI) m/z calcd for C21H29NO8 [M+Na] + 432.1634, found 432.1628;
UPLC TR 1.19 min.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 47
1
2
3
4
((Benzyloxy)carbonyl)ꢀ
D
ꢀglutamic acid (11): Glutamic acid (10g, 68 mmol, 1 equiv) was
5
6
7
8
dissolved in 100 mL DI water at 0 oC (without Argon protection). Sodium carbonate (21.6g, 204
mmol, 3 equiv) and sodium bicarbonate (5.8g, 68 mmol, 1 equiv) were added sequentially. 10mL
acetone was added into the above mixture, followed by benzyl chloroformate (20 mL, 150 mmol,
2.2 equiv). The resulting mixture was stirred for 36 h, gradually warming to room temperature. It
was quenched by acidifying to pH=3 with 20% hydrochloric acid and extracted with EtOAc (3 x
500 mL). The combined organic layers were dried over MgSO4, filtered, and concentrated.
Purification over silica gel chromatography (5% MeOH/CH2Cl2) afforded product 11 as a white
solid (17.5g, 92% yield). [α]ꢁꢀꢄ= +7.2 (c 1.0, MeOH);
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (500 MHz, DMSOꢀd6): δ12.37 (br, 2H), 7.37ꢀ7.28 (m, 5H), 5.02 (s, 2H), 3.98 (td, J
= 11.5, 5.0Hz, 1H), 2.32ꢀ2.26 (m, 2H), 1.98ꢀ1.91 (m, 1H), 1.78ꢀ1.70 (m, 1H);
13C NMR (100 MHz, DMSOꢀd6): δ174.19, 174.06, 156.62, 137.40, 128.79, 128.26, 128.15,
65.87, 53.48, 30.52, 26.52;
IR (neat/cmꢀ1) υmax : 3466, 3332, 3199, 1723, 1666, 1644, 1541, 1410, 1267, 1172, 1051;
HRMS (ESI) m/z calcd for C13H15NO6 [M+Na] + 304.0797, found 304.0789;
UPLC TR 0.82 min.
(R)ꢀ5ꢀaminoꢀ4ꢀ(((benzyloxy)carbonyl)amino)ꢀ5ꢀoxopentanoic acid (12): 11 (15g, 53.3
mmol, 1 equiv) was dissolved in 100mL THF under argon at room temperature. Freshly distilled
triethylamine (16.4mL, 117.3 mmol, 2.2 equiv) was added dropwise into the above solution, and
o
o
stirred for 15 min before moved to a ꢀ30 C chiller. At ꢀ30 C, isobutylchloroformate (8 mL, 64
mmol, 1.0 equiv) was added in dropwise, forming a heterogeneous mixture. It was stirred at ꢀ30
oC for 1h. Ammonia (266mL, 0.5M in THF, 133 mmol, 2.5 equiv) was added in slowly through
the inner surface of the reaction flask. The reaction mixture was stirred at ꢀ30 oC for 48 h before
ACS Paragon Plus Environment

Page 25 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
it was quenched by 200 mL DI water. The mixture was acidified to pH=3 by phosphoric acid.
The resulting mixture was partitioned between 500 mL water and 500 mL ethyl acetate. The
aqueous layer was extracted with EtOAc (3 x 500 mL). The combined organic layers were dried
over MgSO4, filtered, and concentrated. The crude material was used without purification. Pure
product 12 was obtained by preparative thin layer chromatography (TLC) purification in 10%
MeOH/CH2Cl2 for characterization (60% yield). [α]ꢁꢀꢄ= +8.2 (c 1.0, MeOH);
1H NMR (500 MHz, DMSOꢀd6): δ12.07 (br, 1H), 7.35ꢀ7.29 (m, 5H), 7.00 (s, 1H), 5.00 (s,
2H), 3.92 (td, J = 8.8, 5.2Hz, 1H), 2.23 (t, J = 15.7Hz, 2H), 1.91ꢀ1.84 (m, 1H), 1.74ꢀ1.67 (m,
1H);
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13C NMR (100 MHz, DMSOꢀd6): δ174.32, 173.87, 156.38, 137.47, 128.76, 128.21, 128.11,
65.85, 54.31, 30.72, 27.63;
IR (neat/cmꢀ1) υmax : 3465, 3331, 3202, 1723, 1666, 1647, 1609, 1541, 1410, 1352, 1267,
1172, 1052;
HRMS (ESI) m/z calcd for C13H16N2O5 [M+Na] + 303.0957, found 303.0963;
UPLC TR 0.82 min.
(R)ꢀ1ꢀaminoꢀ4ꢀcarboxyꢀ1ꢀoxobutanꢀ2ꢀaminium chloride (13): 12 (12g, 42.8 mmol, 1 equiv)
was dissolved in a mixture of 75 mL methanol and 75 mL acetic acid at room temperature. Pd/C
(10 wt. % loading, 10g, 80 w%) was added in. The internal atmosphere was exchanged with
hydrogen gas by bubbling through hydrogen gas for 5 min. The reaction was stirred under an
atmosphere of hydrogen gas (1 atm, balloon) for 12h. The reaction mixture was then filtered
through a pad of celite and washed with dry methanol. The solvents were removed in vacuo. The
remaining acetic acid was azeotropically removed with ethyl acetate in vacuo (it is important to
make sure that all of the acetic acid was removed). The crude oil was recrystallized by dissolving
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 47
1
2
3
4
5
6
7
8
9
it in methanol and adding concentrated hydrochloric acid to it. The mother liquor was
recrystallized twice until no more solids formed. 13 was obtained as a fluffy, white solid (9g,
62% yield) after air dry.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(R)ꢀ5ꢀaminoꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)ꢀ5ꢀoxopentanoic acid (14): A solution of 13
o
(7.5g, 51.0 mmol, 1 equiv) in 44 mL DI water was cooled to 0 C (without Argon protection).
Sodium carbonate (10.9g, 102.8 mmol, 2 equiv) was slowly added in to the mixture. The
resulting solution was stirred for 10 min until it was clear. 36 mL of 1,4ꢀdioxane was added in,
followed by Diꢀtertꢀbutyl dicarbonate (Boc2O) (13.4g, 61.4 mmol, 1.2 equiv). The mixture was
stirred at room temperature for 12h. It was quenched by phosphoric acid and the pH was adjusted
to 3. The mixture was partitioned between 500 mL ethyl acetate and 500 mL DI water. The
aqueous layer was extracted with EtOAc (3 x 500 mL). The combined organic layers were dried
over MgSO4, filtered, and concentrated. The crude material was used without purification.
Benzyl (R)ꢀ5ꢀaminoꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)ꢀ5ꢀoxopentanoate (15): Crude 14
(12.6g, 51.2 mmol, 1 equiv) was dissolved in 200 mL dry dimethylformamide (DMF) under
argon. Sodium bicarbonate (12.9g, 153.6 mmol, 3 equiv) was added in one portion, followed by
o
benzyl bromide (12 mL, 102.4 mmol, 2 equiv). The mixture was heated at 60 C for 18 h. The
reaction mixture was cooled down to room temperature, and neutralized by phosphoric acid. The
resulting mixture was partitioned between 800 mL ethyl acetate and 500 mL DI water. The
aqueous layer was extracted with EtOAc (3 x 100 mL). The combined organic layers were
washed with 5% LiCl solution (8 x 200 mL) to remove most DMF. Then the organic layer was
dried over MgSO4, filtered, and concentrated. Purification over silica gel chromatography
(gradient, 3% to 4%, MeOH/CH2Cl2) afforded product 15 as a white solid (6.8g, 44% yield over
two steps). [α]ꢁꢀꢄ= +4.6 (c 1.0, CH2Cl2);
ACS Paragon Plus Environment

Page 27 of 47
Journal of Medicinal Chemistry
1
2
1H NMR (500 MHz, CDCl3): δ7.39ꢀ7.30 (m, 5H), 6.30 (br, 1H), 5.65 (br, 1H), 5.33 (br, 1H),
5.13 (s, 2H), 4.21 (br, 1H), 2.57 (td, J = 16.9, 7.4Hz, 1H), 2.46 (td, J = 16.9, 7.0Hz, 1H), 2.20ꢀ
2.13 (m, 1H), 1.97ꢀ1.90 (m, 1H), 1.43 (s, 9H);
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13C NMR (125 MHz, CDCl3): δ174.45, 173.00, 155.82, 135.74, 128.53, 128.23, 128.17,
79.96, 66.46, 53.24, 30.40, 28.31, 27.89;
IR (neat/cmꢀ1) υmax : 3384, 3348, 3184, 2978, 1727, 1682, 1519, 1450, 1391, 1366, 1278, 1250,
1165;
HRMS (ESI) m/z calcd for C17H24N2O5 [M+Na] + 359.1583, found 359.1582;
UPLC TR 1.30 min.
(R)ꢀ1ꢀaminoꢀ5ꢀ(benzyloxy)ꢀ1,5ꢀdioxopentanꢀ2ꢀaminium 2,2,2ꢀtrifluoroacetate (16):
A
solution of 15 (500mg, 1.5 mmol, 1 equiv) in 13 mL dry CH2Cl2 under argon was added in TFA
(1.5 mL, 19.6 mmol, 13 equiv) dropꢀwise. The resulting solution was stirred at room temperature
for 4h. The solvent was removed in vacuo. The crude product 16 was used without purification.
Benzyl (R)ꢀ5ꢀaminoꢀ4ꢀ((S)ꢀ2ꢀ((tertꢀbutoxycarbonyl)amino)propanamido)ꢀ5ꢀoxopentanoate
(17): 16 (525mg, 1.5 mmol, 1 equiv) was taken up in dry dichloromethane and BocꢀLꢀalanine
(227mg, 1.2 mmol, 0.8 equiv) was added under argon. The resulting mixture was cooled down to
o
0 C. TMP (1.3 mL, 7.5 mmol, 5 equiv) was added dropwise into the above mixture. It was
stirred for 5 min before HOAt (306mg, 2.25 mmol, 1.5 equiv) and EDCI (349 mg, 2.25 mmol,
1.5 equiv) were added. The reaction mixture was stirred for 18 h, gradually warming to room
temperature. The reaction was quenched by neutralizing it with saturated ammonia chloride
solution. The mixture was partitioned between 100mL DI water and 100 mL dichloromethane.
The aqueous layer was extracted with dichloromethane (3 x 100mL). The combined organic
layers were dried over MgSO4, filtered, and concentrated. Purification over silica gel
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 47
1
2
3
4
5
6
chromatography (4% MeOH/CH2Cl2) afforded product 17 (403mg, 66% over two steps) as a
fluffy, white solid. [α]ꢁꢀꢄ= +1.4 (c 1.0, CH2Cl2);
7
1H NMR (500 MHz, CDCl3): δ7.38ꢀ7.30 (m, 5H), 7.26 (br, 1H), 6.80 (br, 1H), 5.75 (br, 1H),
5.19 (d, J =3.1, 1H), 5.12 (ABq, J = 12.3 Hz, 2H), 4.48 (dd, J = 10, 7.5Hz, 1H), 4.07, (dq, J =
7.1, 3.1Hz, 1H), 2.57 (dt, J = 16.9, 7.3Hz, 1H), 2.46 (dt, J = 17.0, 6.8 Hz, 1H), 2.26ꢀ2.19, (m,
1H), 2.05ꢀ1.98 (m, 1H), 1.41 (s, 9H), 1.32 (d, J = 7.1Hz, 3H);
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13C NMR (125 MHz, CDCl3): δ173.50, 173.42, 173.32, 155.84, 135.62, 128.57, 128.32,
128.25, 80.38, 66.64, 52.27, 50.64, 30.58, 28.26, 26.77, 17.75;
IR (neat/cmꢀ1) υmax : 3303, 2980, 2936, 1665, 1520, 1455, 1367, 1249, 1164, 1069, 1022;
HRMS (ESI) m/z calcd for C20H29N3O6 [M+Na] + 430.1954, found 430.1951;
UPLC TR 1.25 min.
(S)ꢀ1ꢀ(((R)ꢀ1ꢀaminoꢀ5ꢀ(benzyloxy)ꢀ1,5ꢀdioxopentanꢀ2ꢀyl)amino)ꢀ1ꢀoxopropanꢀ2ꢀaminium
2,2,2ꢀtrifluoroacetate (8): A solution of 17 (400mg, 0.98 mmol, 1 equiv) in 21 mL dry
dichloromethane was added TFA (2.2 mL, 28.8 mmol, 29 equiv) dropwise under argon. The
reaction mixture was stirred at room temperature for 1 h. The solvent was removed in vacuo. The
remaining TFA was azeotropically removed with ethyl acetate in vacuo. The crude product 8
was used without purification.
Benzyl
(R)ꢀ4ꢀ((S)ꢀ2ꢀ((R)ꢀ2ꢀ(((4aR,6S,7R,8R,8aS)ꢀ7ꢀacetamidoꢀ6ꢀ(benzyloxy)ꢀ2,2ꢀ
dimethylhexahydropyrano[3,2ꢀd][1,3]dioxinꢀ8ꢀyl)oxy)propanamido)propanamido)ꢀ5ꢀ
aminoꢀ5ꢀoxopentanoate (4): A solution of 8 (249mg, 0.59 mmol, 1 equiv) in 6 mL
dichloromethane was added 7a (200mg, 0.47 mmol, 0.8 equiv) under argon at 0 oC. To the above
solution, TMP (0.7 mL, 4.13 mmol, 7 equiv) was added dropwise. The resulting solution was
stirred for 5 min. HOAt (120mg, 0.89 mmol, 1.5 equiv), followed by EDCI (138 mg, 0.89 mmol,
ACS Paragon Plus Environment

Page 29 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1.5 equiv) was added sequentially. The reaction mixture was stirred for 18 h, gradually warming
to room temperature. The reaction was quenched by neutralizing it with saturated ammonia
chloride solution. The mixture was partitioned between 100mL DI water and 100 mL
dichloromethane. The aqueous layer was extracted with dichloromethane (3 x 100mL). The
combined organic layers were dried over MgSO4, filtered, and concentrated. Purification over
silica gel chromatography (5% MeOH/CH2Cl2) afforded product 4 (188mg, 56% over two steps)
as a fluffy, white solid. [α]ꢁꢀꢄ= +83.8 (c 1.0, CH2Cl2);
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (500 MHz, CDCl3): δ7.39ꢀ7.30 (m, 10H), 7.15 (d, J = 6.2Hz, 1H), 6.94 (s, 1H), 6.18
(d, J = 9.15Hz, 1H), 5.51 (s, 1H), 5.09 (s, 2H), 4.86 (d, J = 3.9Hz, 1H, anomeric), 4.68 (ABq, J
= 11.9Hz, 1H), 4.48ꢀ4.44 (m, 1H), 4.45 (ABq, J = 11.8Hz), 4.23 (td, J = 9.6, 3.9Hz, 1H), 4.18ꢀ
4.12 (m, 1H), 3.97 (q, J = 6.75Hz, 1H), 3.84ꢀ3.63 (m, 4H), 3.49 (dd, J = 9Hz, 1H), 2.57 (dt, J =
16.8, 7.3 Hz, 1H), 2.46 (dt, J = 17.2, 6.5 Hz, 1H), 2.24ꢀ2.18 (m, 1H), 2.07ꢀ2.00 (m, 1H), 1.92 (s,
3H), 1.50 (s, 3H), 1.40 (d, mix, 3H), 1.39, (s, 3H), 1.35 (d, J = 6.7, 3H);
13C NMR (100 MHz, CDCl3): δ174.36, 173.93, 173.31, 172.10, 170.61, 136.57, 135.47,
128.73, 128.62, 128.45, 128.35, 128.29, 128.06, 99.63, 97.35, 79.20, 78.39, 74.09, 69.96, 66.68,
63.90, 62.27, 52.82, 52.62, 49.94, 30.73, 29.09, 26.29, 23.52, 19.41, 19.04, 16.84;
IR (neat/cmꢀ1) υmax : 3390, 3302, 3055, 2988, 2938, 1736, 1651, 1545, 1455, 1384, 1266, 1203,
1170, 1123, 1077, 1041;
HRMS (ESI) m/z calcd for C36H48N4O11 [M+Na] + 735.3217, found 735.3197;
UPLC TR 1.44 min.
Benzyl
(R)ꢀ4ꢀ((S)ꢀ2ꢀ(2ꢀ(((4aR,6S,7R,8R,8aS)ꢀ7ꢀacetamidoꢀ6ꢀ(benzyloxy)ꢀ2,2ꢀ
dimethylhexahydropyrano[3,2ꢀd][1,3]dioxinꢀ8ꢀyl)oxy)acetamido)propanamido)ꢀ5ꢀaminoꢀ5ꢀ
oxopentanoate (5): A solution of 8 (249mg, 0.59 mmol, 1 equiv) in 6 mL dichloromethane was
ACS Paragon Plus Environment