A highly potent maytansinoid analogue and its use as a cytotoxic therapeutic agent in gold nanoparticles for the treatment of hepatocellular carcinoma

Article abstract of DOI:10.1016/j.bmcl.2020.127634

Gold nanoparticles are promising drug delivery agents with the potential to deliver chemotherapeutic agents to tumour sites. The highly cytotoxic maytansinoid tubulin inhibitor DM1 has been attached to gold nanoparticles and shows tumour growth inhibition in mouse models of hepatocellular carcinoma. Attempting to improve the stability of the gold-cytotoxin bond led to the design and synthesis of novel maytansinoids with improved potency in cell viability assays and improved in vivo tolerability compared to the DM1 analogues. These novel maytansines may also have applications in other methods of drug delivery, for example as the cytotoxic component of antibody drug conjugates.

Full text of DOI:10.1016/j.bmcl.2020.127634

Journal Pre-proofs
A Highly Potent Maytansinoid Analogue and its Use as a Cytotoxic Thera‐
peutic Agent in Gold Nanoparticles for the Treatment of Hepatocellular Carci‐
noma
John Porter, Yao Ding, Sarah J.M. Hale, Richard D. Perrins, Angela
Robinson, Michael P. Mazanetz, Yubo Wu, Yinping Ma, Kelly Conlon, Tom
Coulter
PII:
S0960-894X(20)30745-9
DOI:
Reference:
https://doi.org/10.1016/j.bmcl.2020.127634
BMCL 127634
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
11 May 2020
13 October 2020
17 October 2020
Please cite this article as: Porter, J., Ding, Y., Hale, S.J.M., Perrins, R.D., Robinson, A., Mazanetz, M.P., Wu, Y.,
Ma, Y., Conlon, K., Coulter, T., A Highly Potent Maytansinoid Analogue and its Use as a Cytotoxic Therapeutic
Agent in Gold Nanoparticles for the Treatment of Hepatocellular Carcinoma, Bioorganic & Medicinal Chemistry
Letters (2020), doi: https://doi.org/10.1016/j.bmcl.2020.127634
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. 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.
©
2020 Published by Elsevier Ltd.

A Highly Potent Maytansinoid Analogue and its Use as a Cytotoxic Therapeutic Agent in Gold
Nanoparticles for the Treatment of Hepatocellular Carcinoma
a
a
a
a
a
b
John Porter* , Yao Ding , Sarah JM Hale , Richard D Perrins , Angela Robinson , Michael P Mazanetz ,
c
c
a
a
Yubo Wu , Yinping Ma , Kelly Conlon , Tom Coulter
a
b
Midatech Pharma Ltd., Oddfellows House, 19 Newport Road, Cardiff, CF24 0AA, United Kingdom
NovaData Solutions Ltd., PO Box 639, Abingdon-on-Thames, Oxfordshire, OX14 9JD, United
Kingdom
c
WuXi AppTec (Tianjin) Co., Ltd., No.168 Nan Hai Road, 10th Avenue, TEDA, Tianjin, P.R. China
3
00457
Abstract: Gold nanoparticles are promising drug delivery agents with the potential to deliver
chemotherapeutic agents to tumour sites. The highly cytotoxic maytansinoid tubulin inhibitor DM1
has been attached to gold nanoparticles and shows tumour growth inhibition in mouse models of
hepatocellular carcinoma. Attempting to improve the stability of the gold-cytotoxin bond led to the
design and synthesis of novel maytansinoids with improved potency in cell viability assays and
improved in vivo tolerability compared to the DM1 analogues. These novel maytansines may also
have applications in other methods of drug delivery, for example as the cytotoxic component of
antibody drug conjugates.
Gold nanoparticles (GNPs) have shown promise as drug delivery agents as they are non-toxic, easy to
prepare with controlled size distributions and functionalisable with a range of ligands. The ability to
construct tuneable, multi-modal GNP entities allows the precise control of surface properties for
targeting, stability and, importantly, the release of therapeutic payloads,1,2 and GNPs have been
reported as potential cancer therapies.3-5 One advantage of GNPs is that it is possible to load
relatively large amounts of hydrophobic drugs yet retain good aqueous solubility by the appropriate
use of hydrophilic ligands attached to the gold, effectively changing the in vivo distribution and
pharmacokinetics of the cytotoxic agent. The key requirement is that the active component remains

attached to the gold core until the nanoparticle reaches the target tissue whereupon it is selectively
released in response to the altered physiological state of the diseased (tumour-) microenvironment
compared to healthy tissue. A variety of functional groups will bind to gold, but the gold-sulfur bond
is one of the strongest and is readily formed by the interaction of organic thiols with nanoscale gold.6
Consequently, a thiol based cytotoxic agent would be a valuable tool in the design of a GNP based
delivery system for cancer therapy.
Hepatocellular carcinoma (HCC) is the sixth commonest cancer worldwide with a very poor
7
8,9
prognosis, life expectancy after diagnosis is typically 6-9 months. Although new emerging
technologies such as immuno-oncology are under increasingly active development,10 the current
standard of care for HCC remains surgical resection and/or treatment with sorafenib, a broad-
spectrum kinase inhibitor with a poor tolerability profile and only limited benefit in prolonging
patient survival.11 We have reported our studies on MTC-100038, a GNP conjugated to a cytotoxin
in mouse models of HCC, but it should be noted that without any targeting agents attached to the
gold core, this type of GNP moiety would not be expected to show any tumour selectivity.
12
Maytansine 1 figure 1, an ansa macrolide isolated from Maytenus ovatus,13 is a highly potent
antimitotic agent that exerts an antiproliferative effect by inhibiting microtubule assembly on
1
4
15
binding to tubulin with a K of around 1 μmol/L. Despite a promising in vitro profile, clinical trials
D
with maytansine in cancer patients failed because of poor efficacy and unacceptable systemic
1
6
toxicity. Although the narrow therapeutic window precluded further clinical development work,
maytansines continued to excite interest because of their high cytotoxicity (100 to 1000-fold more
1
7
cytotoxic than the tubulin binders vincristine and vinblastine ). Antibody drug conjugates (ADCs)
utilise monoclonal antibodies that bind to tumour specific receptors with high affinity to provide
both a therapeutic entity, as well as a vehicle for the targeted delivery of chemotherapeutic agents
to cancer cells.18,19 As the antibody concentration at the tumour site is typically low, a highly potent
chemotherapeutic agent is required. Maytansine, with an in vitro IC potency below that of typical
5
0

antibody concentrations at the tumour, has proved an attractive candidate for ADCs. Much effort
has been expended in optimising suitable linkers to conjugate cytotoxins to antibodies,20 for
maytansines this has been achieved by the addition of thiol groups to give DM1, 2, and DM4, 3,
figure 1, with little loss in potency.21a,b The success of this approach may be judged by the FDA
approved Kadcyla®, an ADC consisting of an anti-HER2 antibody conjugated to DM1,22 for the
2
3
treatment of breast cancer, other maytansine based ADCs are progressing through clinical trials. In
some respects cytotoxin conjugated GNPs may be viewed as having some similarities to ADCs, and
potentially have some advantages such as their relative ease of synthesis. However, there remain
significant challenges in the therapeutic development of payload conjugated GNPs, though
promising results for a TNF conjugated GNP in early clinical trials in patients with solid tumours have
been reported.24
O
O
O
O
N
N
SH
Cl
O
1
Cl
O
O
O
O
O
N
O
O
N
O
O
N
O
N
O
HO H
HO H
O
2
O
O
O
SH
N
Cl
O
O
O
O
N
O
N
HO H
O
O
3
Figure 1. Maytansine analogues
DM1 appeared to be an obvious candidate for conjugation to a GNP based on the relative ease of
access, its cytotoxic potency and clinical validation together with the well-known ability of sulfur to
form strong bonds with gold on the nanoscale.25 MTC-100038,12 a 2nm diameter gold core
nanoparticle with galactose and carboxylate oligomeric ethylene glycol passivating and solubilising

ligands, with up to 6 molecules of DM1 per GNP. MTC-100038 was assessed for tolerability in
NOD/SCID mice giving a 3-fold improvement in tolerability compared to DM1 alone, and significant
tumour growth inhibition in an HCC HepB3 murine model bearing subcutaneous xenografts.
However, measuring the rate of cleavage of DM1 from MTC-100038 in biological fluids and
consequent correlation with the observed effects proved challenging (the small quantities,
2
6
metabolism and disulphide bond formation between DM1 and plasma proteins contributing to the
challenges). We therefore sought an alternative method of linking the maytansine moiety to the
gold to simplify analysis and provide improved stability.
Dihydrolipoic acid (DHLA), a dithiol obtained by the reduction of readily available thioctic acid, is an
attractive ligand for addition to GNPs because it adopts a conformation that allows binding by both
sulfur atoms rendering the ligand to be more resistant to displacement by a variety of factors
including pH, ionic strength, heat, light and chemical attack by thiol containing moieties. This
2
7
approach has been shown to improve the colloidal stability of GNPs and resistance to intracellular
2
8
glutathione mediated ligand cleavage. The comparative stability and binding of monothiol, dithiols
2
9
or disulphide ligand binding to GNPs has been reported, however we are not aware of any reports
of the use of this method to link complex cytotoxic molecules to nanoparticles. Once released from
the GNP the disulphide bond would be expected to reform under oxidising conditions, making
covalent binding to plasma proteins less likely, as well as being less prone to metabolism than DM1,
thereby making analysis easier.
The 1,2-dithiolane analogue of DM1 was prepared as shown in Scheme 1. For the initial studies we
used racemic lipoic acid 4 which was coupled to methyl N-methyl-L-alaninate using standard peptide
coupling conditions and the ester hydrolysed with base to give compound 5. Ansamitocin P3 6 was
reduced21a to maytansinol 7 and coupled with compound 5 using DCC and ZnCl to give the desired
2
product 8. The latter step caused the racemisation of the alanine centre, an occurrence that is well-
precedented.21a Although the yields for this step were disappointing, literature reports suggest that a

range of other reagents and conditions have been studied with DCC/ZnCl giving the optimal results.
2
The two products 8 and 9 were separated by preparative tlc, their purity being confirmed by SFC and
HPLC (see supplementary data).
O
O
a,b
HO
HO
N
S
O
S S
S
4
5
O
Cl
O
Cl
O
O
OH
O
O
O
N
O
N
c
d
O
O
N
O
N
O
HO H
HO H
O
O
6
7
O
O
N
O
O
N
Cl
O
8
Cl
O
O
O
S
S
S
O
S
O
N
O
O
N
+
O
O
N
O
N
O
HO H
HO H
O
O
9
Scheme 1: (a) EDCI, HOBT, DIPEA, methyl N-methyl-L-alaninate, DCM, rt; (b) NaOH, H2O, MeOH, rt
o
1
6h; (c) LiAl(OMe) H, THF, -40 C; (d) DCC, ZnCl , compound 5, DCM, rt, 24h
3
2
The cytotoxicity of 8 and 9 was measured by an MTT assay30 against U87MG (glioblastoma) and
Hep3B (hepatocellular carcinoma) cell lines and we were pleasantly surprised to observe that one of
the isomers was significantly more potent than DM1, table 1. This is consistent with the reports for
DM1 analogues where the eutomer was determined to be have the S configuration at the alanine
centre,21a we consequently assigned 8 with S-alanine stereochemistry, though were unable to
confirm this experimentally.
Hep3B cells
SEM
U87MG cells
Alanine
stereo-
chem
1,2-
dithiolane
stereo-chem (nM)
#
IC50
IC fold
IC50
over DM1 (nM)
3.6
IC50 fold
50
n
n
SEM
over DM1
DM1
S
-
5.1 219 0.41
-
171 0.25
-

8
9
S
R
S
R
S
racemic
racemic
0.56 27 0.31
9
0.9
18
0.1
19
0.4
1.2
9.6
0.33
66
0.31
10.6
21 0.69
3
0.4
11
0.05
12
5.8
0.29
76
12 2.02
9
9
9
3
3
1.37
0.03
55.4
0.3
1
0
1
2
3
R
R
S
9
9
3
3
0.01
61
1
1
1
0.27
12.3
0.6
2.5
R
S
1.2
0.3
Table 1: Activity of 1,2-dithiolane analogues in an MTT cytotoxicity assay against Hep3B and U87MG
cells
The potency of 8, 9 and DM1 led us to model their binding to β-tubulin to rationalise the observed
activity and elucidate a possible mechanism of action. Three crystal structures of β-tubulin were
3
1
obtained from the PDB and were used for the construction of a homology model (PDB IDs: 4TUY,
4
TV8, 4TV9). A data set of 25 literature compounds 21a,31 were used as references to establish the
binding site for the docking. Analysis of the results from these docking studies identified several
significant protein-ligand interactions, allowing discrimination between active (pIC values > 9) and
5
0
inactive (pIC values < 9) compounds and the generation of a pharmacophore model to rationalise
5
0
the activities, Figure 1. The sulfur moiety was observed to extend to the GDP binding site, the size of
this feature represents the flexibility in the linker from the amide to the sulfur containing group.
Amino acid residues Phe404 and Trp407 form van der Waals interactions to the alkyl chain of the
ligands. Hydrogen bonds are formed to residues Asn101, Asn102, Lys105, Val181 and Val182. A
feature observed across the chemical series was a pi-pi face-edge interactions with the aryl group of
the ligands and Phe404. The alcohol moiety on the ligands was observed to form either a hydrogen
bond to Ala99 mediated through a water molecule as for ligand 8, or directly to the backbone of
Gly100 as seen with ligand 9. The water-Ala99 mediated hydrogen bond for ligand 8 is in a more
preferential conformation to that formed by either 9 or DM1, this could account for the observed
increase in potency for 8 over these other ligands. An analysis of the contribution to binding energies
showed no significant differences between ligands 8 and 9 to account for the differences in potency,
suggesting that there are other forces driving the protein-ligand interactions. A calculation of ligand

strain energy on both 8 and 9 of the docked poses showed that 8 had a ligand strain energy 12
kcal/mol lower than 9. The energy of solvation (the energy associated with dissolving a ligand in
water, such that a positive number means that dissolution is endothermic whilst a negative figure is
a favourable exothermic process) for 8 and 9 was also calculated. The solvation energy of 8 was -12
kcal/mol lower than that of 9, implying that 9 must undergo a larger conformational change to
become desolvated and bind to β-tubulin than 8. Together these results may explain the observed
differences in potency between 8 and 9.
Figure 2: (A) Structure of the tubulin–DM1 complex. Left; DM1 (cyan) docked to β-tubulin (grey
ribbon), showing significant interactions observed between DM1 and residues of β-tubulin.
Hydrogen bonding interactions are shown as dashed lines. (B) The important pharmacophoric
features for ligand binding to the maytansine site for the docked compounds 8, 9, DM1 (orange,
magenta, cyan sticks respectively). The features are coloured dark blue; included volume required
for the sulphur binding, cyan: HBA group, red; HBA-projection vectors to neighbouring side chains,
green; hydrophobe. Hydrogen bonds shown as dashed lines
These unexpected results prompted us to investigate the effect of resolving the racemic 1,2-
dithiolane centre. Consequently, the S-1,2-dithiolanes 10 and 11, and R-1,2-dithiolanes 12 and 13,

scheme 2, analogues were made in the same way using the homochiral lipoic acid starting material.
However, the eutomer, putative S-alanine, analogues 10 and 12 were only about 2-fold more active
than 8, though interestingly, the R-alanine analogues, 11 and 13, were markedly less potent than 9
in the MTT cytotoxicity assay, Table 1. These results show that binding to the GDP pocket in β-
tubulin is not totally dependent on the 1,2-dithiolane stereochemistry.
O
O
O
O
N
N
Cl
Cl
O
O
O
O
S
S
S
O
S
O
O
N
O
N
O
O
N
O
N
O
HO H
HO H
O
O
10
11
O
O
O
O
N
N
Cl
Cl
O
O
O
O
S
S
S
O
S
O
O
N
O
N
O
O
N
O
N
O
HO H
HO H
O
O
12
13
Scheme 2: Structures of the chiral dithiolane analogues
We next investigated the conjugation of the 1,2-dithiolane cytotoxins to the GNP, scheme 3. The
galactose-carboxylate oligomeric ethylene glycol base nanoparticle was prepared as described for
1
2
32
MTC-100038, scheme 3. Here, a Murray ligand exchange methodology was used to couple the
thiol group of maytansine to the gold core, thereby displacing a galactose or ethylene glycol ligand.
Oh and co-workers were able to displace citrate ligands from GNPs with a disulphide29 but our
attempts were unsuccessful, presumably because of the stronger binding of the galactose- or
ethylene glycol-thiol ligands to the gold core. Consequently, we used more forcing conditions
(addition of TCEP and stirring for 2 hr prior to the addition of the base nanoparticle) to ensure

complete reaction in coupling 8 or 9 to the GNP. This method typically conjugated around 5
molecules of the cytotoxin to each nanoparticle (as determined by HPLC analysis), the resulting dark
brown aqueous solutions were stable when stored at 4°C.
O
HO
H
N
O
O
N
O
Cl
O
O
O
N
O
O
HO
S
OH
HO
S
S
S
HO
O
OH
HO
O
S
O
a
b
O
Au
HO
S
HAuCl
Au
4
O
S
HO
O
COOH
S
8
COOH
S
O
HO
S
8
OH
HO
O
HO
OH
HO
O
8
O
O
8
O
HO
COOH
HO
COOH
Scheme 3: Synthesis of the GNPs (a) (HOOC-PEG(8)-S-) , alpha-GalC , water, NaOH, NaBH ; (b)
2
2
4
Maytansine analogue, TCEP, DMSO, GNP, water
Screening of the S-alanine GNP 14 and the R-alanine GNP 15 in the MTT cytotoxicity assay against
Hep3B and U87MG cell lines showed that 14 retained the relative potency of its non-conjugated
precursor, being 10-fold more potent than its DM1 counterpart (MTC-100038), though the R-alanine
analogue was substantially less potent, table 2. Whereas the stereochemistry of the homochiral 1,2-
dithiolane analogues did not appear to play a significant role, when conjugated to the GNP the S-
alanine, R-1,2-dithiolane 16 was significantly more active, about 30-fold more potent than MTC-
1
00038 and about 3-fold more potent than the S-1,2-dithiolane analogue 18. The reasons for this are
not clear but may involve a more rapid cleavage of the R-1,2-dithiolane from the gold core under the
conditions of the assay.

Hep3B
SEM
U87MG
SEM
1
,2-
Alanine
stereo-
chem
Drug
eq
/GNP (nM)
IC fold
over
MTC-
IC fold
over
MTC-
5
0
50
dithiolane
stereo-
chem
#
IC50
IC50
(nM)
n
n
1
00038
100038
MTC-
S
-
5.5
19.4
9
8.8
-
14.6 18 2.82
1.8 12 0.22
-
1
00038
1
1
1
1
1
1
4
5
6
7
8
9
S
R
S
R
S
racemic
racemic
5.5
5.0
4.3
3.4
5.9
4.8
1.5 12 0.23
13
0.2
33
0.4
6.4
0.6
8
0.2
32
0.5
5.6
0.15
88
0.58
49
2.06
113
3
9
9
3
3
20
68
0.46
30
3
9
9
3
3
13
0.32
9.2
R
R
S
0.38
14.4
0.51
32.6
1.75
83
0.6
2.9
R
S
Table 2: Activity of GNP analogues in an MTT cytotoxicity assay against Hep3B and U87MG cells
The effect of the cytotoxins and cytotoxin conjugated GNPs against a panel of human tumour cancer
cell lines including 786-O (renal adenocarcinoma), A2780 (ovarian cancer), A375 (melanoma),
A431(epidermoid carcinoma), A549 (adenocarcinoma), ACHN (renal adenocarcinoma), BXPC-3
(pancreas adenocarcinoma), as well as Hep3B and U87MG used in the above study. The cell lines
showed varying susceptibility to the unconjugated cytotoxins 8 and 10, but these molecules were at
least an order of magnitude more potent than DM1. The difference between the conjugated
cytotoxin GNPs 14 and 16 and the DM1 conjugated GNP, MTC-100038, were perhaps less profound
but were still significant (See Supplementary Tables 3 and 4). Note that, as yet, we have been unable
to determine the rate of cleavage of the cytotoxin from the GNP and that the predicted slower rate
of cleavage for the 1,2-dithiolane maytansines may well have an effect on the observed cytotoxicity.
Finally, we conducted a tolerability study of maytansine 8 and GNP 14 in mice to compare to DM1
1
2
and MTC-100038. In the previous study, DM1 and MTC-100038 had been exposed to a 2 x QD5 iv
regime, but for this study the exposure was reduced to a weekly iv dose, which was felt to be more
in keeping with a clinical dosing schedule. Tolerability was assessed by change in body weight. Under
these conditions a DM1 dose of 350 ug/kg QW x 3 or 450 ug/kg QW x 2 was well tolerated but
increasing to 1350 ug/kg QW x 1 saw a rapid loss in body weight, Figure 3A. In contrast, the

maximum feasible dose of 8 that could be administered, because of solubility issues, was 1350
ug/kg, this dose was well tolerated, Figure 3B. For the GNP conjugates, MTC-100038 was tolerated
to 1250 ug/kg (drug equivalent dose), Figure 3C, whereas GNP 14 was tolerated to a dose of 2700
ug/kg (drug equivalent dose), Figure 3D, over twice that of MTC-100038, demonstrating that the
strategy of stabilising the linkage of the cytotoxic moiety to the gold core appeared to have a
significant effect in improved tolerability.
In conclusion, we have attempted to improve the properties of a cytotoxic maytansine conjugated to
a GNP by improving the stability of the drug-gold linkage. These modifications had the unexpected
benefit of improving the potency as determined by a cytotoxicity assay against a number of cancer
cell lines. The improvements in potency could be accounted for by computational modelling. The
resulting cytotoxic GNP was significantly better tolerated in a mouse study. These results are
sufficiently promising to warrant further in vivo pharmacokinetic and biodistribution studies as well
as efficacy studies in mouse models of HCC. Additionally, we need to understand the rate of
cleavage of the cytotoxin to help interpret these data. We have focused on the use of these novel
cytotoxic maytansines in combination with gold nanoparticles, but it is apparent to the authors that
these agents may have applications in other areas such as antibody drug conjugate systems where
their potency and the unusual method of attachment through the dithiolane group may offer
advantages.

Figure 3. Tolerability in mice dosing QW. A. DM1 tolerability, B. Maytansinoid 8 tolerability, C. GNP
MTC-100038 tolerability, D. GNP 14 tolerability
References and Notes
1
2
.
.
Yang, X.; Yang, M.; Pang, B.; Vara, M.; Xia, Y.; Gold Nanomaterials at Work in Biomedicine.
Chemical Reviews 2015, 115 (19), 10410-10488, doi: 10.1021/acs.chemrev.5b00193
Kumar, A.; Zhang, X.; Liang, X. J., Gold nanoparticles: emerging paradigm for targeted drug
delivery system. Biotechnology advances 2013, 31 (5), 593-606, doi:
1
0.1016/j.biotechadv.2012.10.002

3
4
.
.
Sharma, A.; Goyal, A. K.; Rath, G., Recent advances in metal nanoparticles in cancer therapy.
Journal of Drug Targeting 2018, 26 (8), 617-632, doi: 10.1080/1061186X.2017.1400553
F. Jiao, P.; Y. Zhou, H.; X. Chen, L.; Yan, B.; Cancer-Targeting Multifunctionalized Gold
Nanoparticles in Imaging and Therapy. Current Medicinal Chemistry 2011, 18 (14), 2086-
2
102, doi: 10.2174/092986711795656199
5
6
7
.
.
.
Li, H.; Jin, H.; Wan, W.; Wu, C.; Wei, L.;, Cancer nanomedicine: mechanisms, obstacles and
strategies. Nanomedicine 2018, 13 (13), 1639-1656, doi: 10.2217/nnm-2018-0007
Hakkinen, H., The gold-sulfur interface at the nanoscale. Nat Chem 2012, 4 (6), 443-455, doi:
1
0.1038/nchem.1352
McGlynn, K. A.; Petrick, J. L.; London, W. T., Global epidemiology of hepatocellular
carcinoma: an emphasis on demographic and regional variability. Clinics in liver disease
2
015, 19 (2), 223-238, doi: 10.1016/j.cld.2015.01.001
8
.
Ferlay, J.; Soerjomataram, I.; Dikshit, R.; Eser, S.; Mathers, C.; Rebelo, M.; Parkin, D.M;
Forman, D.; Bray, F.; Cancer incidence and mortality worldwide: sources, methods and
major patterns in GLOBOCAN 2012. Int J Cancer 2015, 136 (5), E359-86, doi:
1
0.1002/ijc.29210
Parkin, D. M.; Bray, F.; Ferlay, J.; Pisani, P.; Estimating the world cancer burden: Globocan
000. Int J Cancer 2001, 94 (2), 153-6, doi: 10.1002/ijc.1440.
9
1
.
2
0.
Topalian, S. L.; Drake, C. G.; Pardoll, D. M., Immune checkpoint blockade: a common
denominator approach to cancer therapy. Cancer Cell 2015, 27 (4), 450-61, doi:
1
0.1016/j.ccell.2015.03.001
1
1
1.
2.
Zhong, J. H.; Du, X. K.; Xiang, B. D. Li, L-Q.; Adjuvant sorafenib in hepatocellular carcinoma: A
cautionary comment of STORM trial. World J Hepatol 2016, 8 (23), 957-60, doi:
1
0.4254/wjh.v8.i23.957
Hale, S. J. M.; Perrins, R. D.; Garcı
Williams, P.; Ding, Y.; Saito, G.; Rodriguez, M.A.; Perera, I.; Barrientos, A.; Conlon, K.;
́a , C. E. ; Pace, A.; Peral, U.; Patel, K.R.; Robinson, A.;

Damment, S.; Porter, J.; Coulter, T., DM1 Loaded Ultrasmall Gold Nanoparticles Display
Significant Efficacy and Improved Tolerability in Murine Models of Hepatocellular Carcinoma.
Bioconjugate Chemistry 2019, 30 (3), 703-713, doi 10.1021/acs.bioconjchem.8b00873
Kupchan, S. M.; Komoda, Y.; Court, W. A.; Thomas, G.J.; Smith R.M.; Karim, A.; Gilmore, C.J.;
Haltiwanger, R.C.; Bryan, R.F.;, Tumor inhibitors. LXXIII. Maytansine, a novel antileukemic
ansa macrolide from Maytenus ovatus. Journal of the American Chemical Society 1972, 94
1
1
1
3.
4.
5.
(4), 1354-1356, doi: 10.1021/ja00759a054
Lopus, M.; Oroudjev, E.; Wilson, L.; Wilhelm, S.; Widdison, W.; Chari, R.; Jordan, M.A.;
Maytansine and cellular metabolites of antibody-maytansinoid conjugates strongly suppress
microtubule dynamics by binding to microtubules. Mol Cancer Ther 2010, 9 (10), 2689-99,
doi: 10.1158/1535-7163.MCT-10-0644
Cassady, J. M.; Chan, K. K.; Floss, H. G.; Leistner, E.; Recent developments in the
maytansinoid antitumor agents. Chemical & Pharmaceutical bulletin 2004, 52 (1), 1-26, doi:
1
0.1248/cpb.52.1
Issell, B. F.; Crooke, S. T., Maytansine. Cancer Treatment Reviews 1978, 5 (4), 199-207, doi:
0.1016/S0305-7372(78)80014-0
1
1
6.
7.
1
Kupchan, S. M.; Komoda, Y.; Branfman, A. R.; Sneden, A.T.; Court, W.A.; Thomas, G.J.; Hintz,
H.P.J.; Smith, R.M.; Karim A.; Howie, G.A.; Verma, A.K.; Nagao, Y.; Dailey, R.G.; Zimmerly,
V.A.; Sumner Jr., W.A.; Tumor inhibitors. 122. The maytansinoids. Isolation, structural
elucidation, and chemical interrelation of novel ansa macrolides. The Journal of Organic
Chemistry 1977, 42 (14), 2349-2357, doi: 10.1021/jo00434a001
1
8.
Sau, S.; Alsaab, H. O.; Kashaw, S. K.; Tatiparti, K.; Iyer, A. K., Advances in antibody–drug
conjugates: A new era of targeted cancer therapy. Drug Discovery Today 2017, 22 (10),1547-
1
556, doi: 10.1016/j.drudis.2017.05.011

1
9.
Kumar, A.; White, J.; Christie, J.R.; Dimasi, N.; Gao, C., Antibody-Drug Conjugates. In Annual
Reports in Medicinal Chemistry, Goodnow, R. A., Ed. Academic Press: 2017, 50, pp 441-480,
doi: 10.1016/bs.armc.2017.08.002
2
2
0.
1.
Jain, N.; Smith, S. W.; Ghone, S.; Tomczuk, B., Current ADC Linker Chemistry. Pharmaceutical
Research 2015, 32 (11), 3526-3540, doi: 10.1007/s11095-015-1657-7
a: Widdison, W. C.; Wilhelm, S. D.; Cavanagh, E.E.; Whiteman, K.R.; Leece, B.A.; Kovtun, Y.;
Goldmacher, V.S.; Xie, H,; Steeves, R.M.; Lutz, R.J.; Zhao, R.; Wang, L.; Blattler, W.A.; Chari,
R.V.J., Semisynthetic Maytansine Analogues for the Targeted Treatment of Cancer. Journal
of Medicinal Chemistry 2006, 49 (14), 4392-4408, doi: 10.1021/jm060319f; b: Chen, H., Lin,
Z., Arnst K.E., Miller D.D., Li, W., Tubulin Inhibitor-Based Antibody-Drug Conjugates for
Cancer Therapy Molecules 2017, 22, 1281, doi: 10.3390/molecules22081281 and references
therein.
2
2
2
2.
3.
4.
Lambert, J. M.; Morris, C. Q., Antibody-Drug Conjugates (ADCs) for Personalized Treatment
of Solid Tumors: A Review. Advances in therapy 2017, 34 (5), 1015-1035, doi:
1
0.1007/s12325-017-0519-6
Beck, A.; Goetsch, L.; Dumontet, C.; Corvaia, N., Strategies and challenges for the next
generation of antibody-drug conjugates. Nat Rev Drug Discov 2017, 16 (5), 315-337, doi:
1
0.1038/nrd.2016.268
Libutti, S. K.; Paciotti, G. F.; Byrnes, A. A.; Alexander, H. R. Jr.; Gannon, W. E.; Walker, M.;
Seidel, G. D.; Yuldasheva, N.; Tamarkin, L. Phase I and Pharmacokinetic Studies of CYT-6091,
a Novel Pegylated Colloidal Gold-rh TNF Nanomedicine. Clin. Cancer Res. 2010, 16,
6
139−6149, doi: 10.1158/1078-0432.CCR-10-0978
2
5.
Chi, Q.; Ford, M. J.; Halder, A.; Hush, N.S.; Reimers, J.R.; Ulstrup, J., Sulfur ligand mediated
electrochemistry of gold surfaces and nanoparticles: What, how, and why. Current Opinion in
Electrochemistry 2017, 1 (1), 7-15, doi: 10.1016/j.coelec.2016.12.004

2
6.
Shen, B.-Q.; Bumbaca, D.; Yue, Q.; Saad, O.; Tibbitts, J; Khojasteh, S.C.; Girish, S., Non-Clinical
Disposition and Metabolism of DM1, a Component of Trastuzumab Emtansine (T-DM1), in
Sprague Dawley Rats. Drug Metabolism Letters 2015, 9 (2), 119-131, doi:
1
0.2174/1872312809666150602151922
2
2
2
7.
8.
9.
Mei, B. C.; Oh, E.; Susumu, K.; Farrell, D.; Mountziaris, T.J.; Mattoussi, H., Effects of Ligand
Coordination Number and Surface Curvature on the Stability of Gold Nanoparticles in
Aqueous Solutions. Langmuir 2009, 25 (18), 10604-10611, doi: 10.1021/la901423z
Zhu, Z. J.; Tang, R.; Yeh, Y. C.; Miranda, O.R.; Rotello, V.M.; Vachet, R.W., Determination of
the intracellular stability of gold nanoparticle monolayers using mass spectrometry. Anal
Chem 2012, 84 (10), 4321-6, doi: 10.1021/ac203408v
Oh, E.; Susumu, K.; Mäkinen, A. J.; Deschamps, J. R.; Deschamps, J.R.; Huston, A.L.; Medintz,
I.L., Colloidal Stability of Gold Nanoparticles Coated with Multithiol-Poly(ethylene glycol)
Ligands: Importance of Structural Constraints of the Sulfur Anchoring Groups. The Journal of
Physical Chemistry C 2013, 117 (37), 18947-18956, doi: 10.1021/jp405265u
Patravale, V.; Dandekar, P.; Jain, R., Nanotoxicology: evaluating toxicity potential of drug-
nanoparticles. In Nanoparticulate Drug Delivery, Patravale, V.; Dandekar, P.; Jain, R., Eds.
Woodhead Publishing: 2012; pp 123-155, doi: 10.1533/9781908818195.123
Prota, A.E., Bargsten, K., Diaz, J.F.; Marsh, M.; Cuevas, C.; Liniger, M.; Neuhaus, C.; Andreu,
J.M.; Altmann, K-H.; Steinmetzet, M.O., A new tubulin-binding site and pharmacophore for
microtubule-destabilizing anticancer drugs, Proc Natl Acad Sci USA, 2014, 111, 13817-13821,
doi: 10.1073/pnas.1408124111
3
3
0.
1.
3
2.
Templeton, A. C.; Wuelfing, W. P.; Murray, R. W., Monolayer-Protected Cluster Molecules.
Accounts of Chemical Research 2000, 33 (1), 27-36, 10.1021/ar9602664

Figure and scheme captions
Figure 1. Maytansine Analogues
Figure 2: (A) Structure of the tubulin–DM1 complex. Left; DM1 (cyan) docked to β-tubulin (grey
ribbon), showing significant interactions observed between DM1 and residues of β-tubulin.
Hydrogen bonding interactions are shown as dashed lines. (B) The important pharmacophoric
features for ligand binding to the maytansine site for the docked compounds 8, 9, DM1 (orange,
magenta, cyan sticks respectively). The features are coloured dark blue; included volume required
for the sulphur binding, cyan: HBA group, red; HBA-projection vectors to neighbouring side chains,
green; hydrophobe. Hydrogen bonds shown as dashed lines
Figure 3. Tolerability in mice dosing QW. (A) DM1 tolerability, (B) Maytansinoid 8 tolerability, (C)
GNP MTC-100038 tolerability, (D) GNP 14 tolerability
Table 1: Activity of 1,2-dithiolane analogues in an MTT cytotoxicity assay against Hep3B and U87MG
cells
Table 2: Activity of GNP analogues in an MTT cytotoxicity assay against Hep3B and U87MG cells
Scheme 1: (a) EDCI, HOBT, DIPEA, methyl N-methyl-L-alaninate, DCM, rt; (b) NaOH, H2O, MeOH, rt
1
6h; (c) LiAl(OMe)3H, THF, -40oC; (d) DCC, ZnCl2, compound 5, DCM, rt, 24h
Scheme 2: Structures of the chiral dithiolane analogues

Scheme 3: Synthesis of the GNPs (a) (HOOC-PEG(8)-S-)2, alpha-GalC2, water, NaOH, NaBH4; (b)
Maytansine analogue, TCEP, DMSO, GNP, water

Declaration of interests
☒
The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
☐
The authors declare the following financial interests/personal relationships which may be
considered as potential competing interests:
\