Synthesis of N-Hydroxy Isopeptide Containing Proteins

Article abstract of DOI:10.1021/acs.orglett.5b01443

Chemical synthesis of a peptide-ubiquitin conjugate linked by an N-hydroxy isopeptide bond to determine what effect the N-hydroxy group has on the enzymatic hydrolysis of the isopeptide linkage by deubiquitinases is reported. This conjugate was subjected

Full text of DOI:10.1021/acs.orglett.5b01443

Letter
pubs.acs.org/OrgLett
Synthesis of N-Hydroxy Isopeptide Containing Proteins
Suman Kumar Maity, Shimrit Ohayon, and Ashraf Brik*
Schulich Faculty of Chemistry TechnionIsrael Institute of Technology, 3200008 Haifa, Israel
S
* Supporting Information
ABSTRACT: Chemical synthesis of a peptide−ubiquitin
conjugate linked by an N-hydroxy isopeptide bond to determine
what effect the N-hydroxy group has on the enzymatic
hydrolysis of the isopeptide linkage by deubiquitinases is
reported. This conjugate was subjected to proteolysis by
UCH-L3 in the presence and absence of various metal ions,
and no substantive difference in hydrolysis was seen compared
to a control lacking the N-hydroxy group. The accessibility of N-
hydroxy ubiquitinated substrates may find uses to study other
deubiquitinases in particular those which use a zinc ion as a part
of their catalytic mechanism.
he ability to fine-tune the atomic structure of peptides and
Tproteins enables researchers to introduce new properties
into these macromolecules that are important in various
research areas. Compared to proteins, peptides and peptidomi-
metics are more amenable for such modifications because of
their smaller size and relatively straightforward synthesis. For
example, peptidomimetics such as siderophores, which contain
N-hydroxy amide bonds as chelators of Fe(III) ions to assist
cell membrane transport in microorganisms,1 have been
synthesized and extensively studied. N-Hydroxy amide
containing peptides and peptidomimetics, which exhibit
excellent inhibition properties for metalloproteases such as
thermolysin,2 methionine aminopeptidase,3 leukotriene A4
hydrolase,4 encephalin degrading enzyme,5 and peptide
deformylase,6 have also been synthesized and studied.
However, to the best of our knowledge, the synthesis of
proteins containing N-hydroxy amide has not been reported so
far. Here we report on the synthesis of ubiquitinated peptides
bearing a N-hydroxy isopeptide bond to examine the effect of
this modification on the cleavage of this bond by known
proteases.
It has been reported that the N-hydroxy amide inserted in
short peptides exhibits different stability to enzymatic
hydrolysis compared to the native amide.7 We reasoned that
the introduction of this moiety into the isopeptide bond linking
peptides or proteins through the Lys side chain to the C-
terminal of ubiquitin or ubiquitin-like modifiers (e.g., SUMO)
could in principle influence the electronic and conformational8
properties of the isopeptide bond. As a result this might affect
the enzymatic hydrolysis by deubiquitinases (DUBs), which are
known to cleave the isopeptide bond.9 To test this hypothesis
we set the goal of preparing a short peptide linked to ubiquitin
(Figure 1), applying state-of-the-art chemical protein synthesis
approaches.
Figure 1. Schematic presentation of a short peptide linked to protein
(e.g., ubiquitin) through (A) native and (B) N-hydroxy isopeptide
bonds.
Scheme S1). For example, acylation of free amine of a
tripeptide by N-benzyloxy-N-carboxy α-amino acid anhydrides
afforded the corresponding N-benzyloxytetrapeptide.10 N-
Hydroxy glycine residue was incorporated in the peptide
backbone by coupling of bromoacetic acid on a solid support
followed by nucleophilic substitution by O-benzylhydroxyl-
amine and subsequent peptide elongation.7a In another report,
it was found that acylation of N-(benzoyloxy) phenethylamine
with the acid chloride of Nα-Fmoc-L-leucine provided an Nα-
Fmoc-N-(benzoyloxy)-L-leucinamide, where the free hydroxyl
can be achieved by benzoyl group deprotection.11 Coupling of
α-amino acids with α-keto acyl chlorides was found to produce
the corresponding α-keto amides, which upon condensation
with hydroxylamine gave the oxime bond. It has also been
reported that reduction of oxime followed by selective N-
acylation with Fmoc-protected amino acid chlorides afforded
the N-hydroxy peptide linkage.12 Bode and co-workers reported
the synthesis of enantiopure 5-oxaproline and N-benzylidene
nitrone protected N-hydroxyamino acids, which were readily
incorporated in peptides by solid phase peptide synthesis
(SPPS) and used in chemical protein synthesis.13 In addition,
the ε-amine of lysine has been hydroxylated by oxidation with
dibenzoyl peroxide followed by acylation with formic acid.14
Several methods have been reported for the synthesis of N-
hydroxy amide containing peptides (Supporting Information,
Received: May 17, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b01443
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Despite the fact that these methods have been successfully
exercised on peptides, their extension to proteins is not trivial.
In this report, we present the first synthesis of a protein
containing a N-hydroxy isopeptide bond. We chose ubiquitin as
a protein model since the preparation of ubiquitinated peptides
with a more labile isopeptide bond might enable the synthesis
of more efficient substrates for high throughput screening of
inhibitors against DUBs.15 Moreover, since some DUBs employ
a zinc ion as a part of their catalytic apparatus, such conjugates
could also be used to trap the catalytic zinc and generate
inactive DUB bound to its substrate for a variety of studies. To
achieve the preparation of such a conjugate, we synthesized Lys
derivative 1a starting from Nα-Boc-Lys-OMe by oxidation of
the ε-amine with dibenzoyl peroxide to N-(benzoyloxy) amine
followed by acylation with Fmoc-Gly-OH using 1-[Bis-
(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]-
pyridinium 3-oxid hexafluorophosphate (HATU) as a coupling
agent (Scheme 1). This derivative was incorporated on a Rink
amide resin via conventional Fmoc-SPPS (Scheme 1).
However, in the case of 1a, we did not observe further
coupling on the free amine (Supporting Information, Figure
S1).
We then wondered if SPPS would proceed without particular
problems subsequent to the incorporation of 1b, which has two
glycine residues at the ε-amine of the Lys residue. In this case,
despite the fact that we were able to couple the next amino
acids, several unidentified side products were observed as we
progressed with the synthesis (Supporting Information, Figures
S2 and S3). Moreover, when coupling was performed after 48 h
of the Fmoc removal, the majority of the starting material
remained unreacted (Supporting Information, Figure S4). We
reasoned that the amine was capped by the transfer of the
benzoyl protecting group from the N-benzoyloxy group to the
glycine amine, after the piperidine treatment step (Supporting
Information, Scheme S2).16
We then changed the benzoyl-protecting group to the benzyl
derivative and synthesized the modified Lys with two glycine
residues (Scheme 1). The incorporation of a benzyl protecting
group was carried out through solvolysis in methanol and N,N-
diisopropylethylamine (DIEA) in the presence of benzyl
bromide (Scheme 1). Compound 2a was successfully
incorporated on the resin and the couplings of the next
amino acids were performed smoothly, supporting that the free
amine was capped in the case of benzoyl protection. With this
approach, we synthesized the Cys-Ub(47−76) fragment with a
mutation of Ala46Cys to enable native chemical ligation
(NCL)17 with the complementary Ub(1−45)-thioester. The
Nα-Alloc protecting group on the backbone peptide was
removed followed by coupling of Ser and Leu. Unfortunately,
the removal of the benzyl-protecting group using either
hydrogenation over Pd/C or trifluoromethanesulfonic acid
(TFMSA) was problematic and a major product with a mass
that was 16 Da less than that of the desired product was
obtained (Supporting Information, Figure S5). This probably
occurred due to the removal of an oxygen atom from the N-
benzyloxy functionality during the deprotection step.
Scheme 1. Synthesis of Peptide Fragment Containing N-
Hydroxy Isopeptide from Lysine Derivatives
At this point we thought to divide the Cys-Ub(47−76)
fragment into two fragments, which could lead to a more facile
deprotection. Accordingly, the Cys-Ub(47−76) peptide was
divided into two fragments, one containing C-terminal N-acyl-
benzimidazolinone (Nbz) as well as N-terminal thiazolidine
(Thz), Thz-Ub(47−66)-Nbz (Supporting Information, Figure
S7), while the other contains the N-hydroxy isopeptide bond
with N-terminal Cys, Cys-Ub-(68−76), to enable NCL.
Fortunately, by using this approach the benzyl removal was
successfully achieved employing a mixture of TFMSA/TFA/m-
cresol (Supporting Information, Figure S6 and Scheme S3).
Subsequently, the short peptides were ligated and the N-
terminal Thz was unmasked by methoxylamine to produce Cys-
Ub(47−76), ready for a second ligation (Scheme 2). This
fragment was then ligated with Ub(1−45)-thioester to produce
the full ubiquitin with the N-hydroxy isopeptide bond
(Supporting Information, Figures S8 and S9). The ubiquiti-
nated peptide with the native isopeptide bond, as a control, was
also prepared from similar fragments (Supporting Information,
Figures S10 and S11).
To check the effect of N-hydroxy functionality on the
electronic environment of the isopeptide carbonyl bond, we
synthesized the LSKAKE peptide linked to the Ub(72−76)
fragment with and without the N-hydroxy functionality, where
the Gly76 residue contained 13C enriched carbonyl (Supporting
Information, Figures S12 and S13). The carbonyl from the N-
B
DOI: 10.1021/acs.orglett.5b01443
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Synthesis of N-Hydroxy Ubiquitinated Peptide
from Three Fragments; HPLC and Mass Spectra of the Pure
N-Hydroxy Ubiquitinated Peptide with the Observed
Average Mass 9298.8 0.6 Da, Calcd 9297.7 Da (Average
Isotopes)
Figure 2. 13C NMR spectra of (a) Ub(72−76)-peptide (b) N-
hydroxy-Ub(72−76)-peptide with 13C labeled isopeptide bond with
and without Zn(II) ions. (*) is the reference peak from methanol-d4.
(c) The extent of hydrolysis of ubiquitinated and N-hydroxy
ubiquitinated peptides with respect to different equivalents of Zn(II)
ions by UCH-L3.
hydroxy isopeptide exhibited a δ value of 170.54 ppm, whereas
that of the native isopeptide exhibited 172.06 ppm (Figure 2),
indicating that the carbon is slightly electron rich compared to
the native amide carbon. Next, we started to examine their
behavior with known DUBs, and we initially chose to work with
UCH-L3 since it is known to cleave small peptides from the C-
terminal of ubiquitin. The cleavage reactions were monitored
by HPLC following the ubiquitin formation. Testing both
substrates revealed that the extent of hydrolysis of the
ubiquitinated peptide bearing an N-hydroxy isopeptide bond
was lower by only 8% compared to the native isopeptide bond
(Figure 2C).
Next, we wondered about effect of metal ions binding on the
electronic environment of the amide carbonyl and the
enzymatic hydrolysis. The 13C NMR of the N-hydroxy in the
presence of 1 equiv of Zn(II) ion in HEPES buffer showed a
large upfield shift in the δ value, from 170.54 to 161.78 ppm,
whereas the native isopeptide bond did not show any shift
(Figure 2a and 2b).18 This indicates that due to metal ion
binding the carbonyl group of the N-hydroxy isopeptide bond
becomes more electron rich.
addition of 5 equiv of Zn(II), and only with an increasing
amount of Zn(II) ions we observed an additional decrease in
the hydrolysis in the case of the N-hydroxy ubiquitinated
peptide (Figure 2c). This may suggest that Zn(II) ions also
bind to ubiquitin (as it contains cysteine and histidine residues)
and/or to the enzyme. The excess of Zn(II) ions may harm the
activity of the enzyme, as we observed in the case of the
ubiquitinated peptide where the extent of hydrolysis was
decreased with an increasing amount of zinc. In the case of the
N-hydroxy ubiquitinted peptide despite the relatively large
chemical shift in the NMR upon addition of Zn(II) ions, the
extent of hydrolysis was lower only by 13%.19
We also studied the effect of Fe(III) ions on the hydrolysis of
ubiquitinated and N-hydroxy ubiquitinated peptides by UCH-
L3, where we observed that 15 equiv of Fe(III) ions had no
effect on the hydrolysis (Supporting Information, Figure S16).
The 13C NMR of the labeled peptides in the presence of
Ga(III) ions which show similar coordination properties such
as Fe(III) ions20 exhibited no shift in the 13C spectra of the N-
hydroxy model peptide suggesting that Ga(III) ions were not
coordinating with the N-hydroxy functionality (Supporting
We next studied the effect of Zn(II) ions on the hydrolysis of
ubiquitinated peptides containing N-hydroxy and native
isopeptide bonds by UCH-L3. Surprisingly, with UCH-L3, no
difference was observed in the extent of hydrolysis with the
C
DOI: 10.1021/acs.orglett.5b01443
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(11) Wang, L.; Phanstiel, O. J. Org. Chem. 2000, 65, 1442.
(12) Lawrence, J.; Cointeaux, L.; Maire, P.; Vallee, Y.; Blandin, V.
Org. Biomol. Chem. 2006, 4, 3125.
(13) (a) Pattabiraman, V. R.; Ogunkoya, A. O.; Bode, J. W. Angew.
Chem., Int. Ed. 2012, 51, 5114. (b) Medina, S. I.; Wu, J. A.; Bode, J. W.
Org. Biomol. Chem. 2010, 8, 3405.
(14) Fennell, K. A.; Miller, M. J. Org. Lett. 2007, 9, 1683.
(15) (a) Ohayon, S.; Spasser, L.; Aharoni, A.; Brik, A. J. Am. Chem.
Soc. 2012, 134, 3281. (b) Ohayon, S.; Refua, M.; Hendler, A.; Aharoni,
A.; Brik, A. Angew. Chem., Int. Ed. 2014, 54, 599.
(16) Braslau, R.; Axon, J. R.; Lee, B. Org. Lett. 2000, 2, 1399.
(17) Dawson, P. E.; Muir, T. W.; Clarklewis, I.; Kent, S. B. H. Science
1994, 266, 776.
(18) Zoroddu, M. A.; Medici, S.; Peana, M.; Anedda, R. Dalton Trans.
2010, 39, 1282.
Information, Figure S17). Finally, we examined the effect of
hydrolysis by UCH-L3 in the presence of Cu(II) ions.21 In
presence of 15 equiv of Cu(II) ions, the ubiquitinated peptide
was hydrolyzed to 58% and the N-hydroxy ubiquitinated
peptide was hydrolyzed to 45% (Supporting Information,
Figure S18). This indicated that Cu(II) ions had less effect on
ubiquitin and/or enzyme compared to Zn(II) ions, as the
extent of hydrolysis with Cu(II) ions was significantly more,
compared to the presence of 15 equiv of Zn(II) ions
(Supporting Information, Table S1).
In summary, for the first time the synthesis of a N-hydroxy
isopeptide bond-containing protein was achieved by applying
small molecule, peptide, and protein synthesis. The presence of
such functionality in the free or metal bound forms did not
significantly affect the UCH-L3 activity. Expanding this study to
other DUBs is currently underway in our laboratory.
(19) Bozza, W. P.; Liang, Q.; Gong, P.; Zhuang, Z. Biochemistry 2012,
51, 10075. Note: See this reference where the attack of the catalytic
cysteine on the isopeptide carbonyl is not shown to be the rate-
determining step in the DUB mechanism.
(20) Hara, Y.; Shen, L. T.; Tsubouchi, A.; Akiyama, M.; Umemoto,
K. Inorg. Chem. 2000, 39, 5074.
ASSOCIATED CONTENT
* Supporting Information
■
S
(21) Farkas, E.; Kiss, I. J. Chem. Soc., Dalton Trans. 1990, 749.
All experimental procedures, analytical data of synthetic
compounds, enzymatic hydrolysis studies of ubiquitinated
peptides. The Supporting Information is available free of
charge on the ACS Publications website at DOI: 10.1021/
acs.orglett.5b01443.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: abrik@technion.ac.il.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Israel Science Foundation for financial support
(A.B.). S.K.M. thanks the Israel Council of Higher Education
for a fellowship under the PBC program. A.B. is a Neubauer
Professor and a Taub Fellow-Supported by the Taub
Foundations.
REFERENCES
■
(1) Drechsel, H.; Jung, G. J. Pept. Sci. 1998, 4, 147.
(2) Nishino, N.; Powers, J. C. Biochemistry 1978, 17, 2846.
(3) Hu, X. B.; Zhu, J. G.; Srivathsan, S.; Pei, D. H. Bioorg. Med. Chem.
Lett. 2004, 14, 77.
(4) Hogg, J. H.; Ollmann, I. R.; Wetterholm, A.; Andberg, M. B.;
Haeggstrom, J.; Samuelsson, B.; Wong, C. H. Chem.Eur. J. 1998, 4,
1698.
(5) Bouboutou, R.; Waksman, G.; Devin, J.; Fourniezaluski, M. C.;
Roques, B. P. Life Sci. 1984, 35, 1023.
(6) Apfel, C.; Banner, D. W.; Bur, D.; Dietz, M.; Hubschwerlen, C.;
Locher, H.; Marlin, F.; Masciadri, R.; Pirson, W.; Stalder, H. J. Med.
Chem. 2001, 44, 1847.
(7) (a) Bianco, A.; Zabel, C.; Walden, P.; Jung, G. J. Pept. Sci. 1998, 4,
471. (b) Bianco, A.; Kaiser, D.; Jung, G. J. Pept. Res. 1999, 54, 544.
(8) Marraud, M.; Dupont, V.; Grand, V.; Zerkout, S.; Lecoq, A.;
Boussard, G.; Vidal, J.; Collet, A.; Aubry, A. Biopolymers 1993, 33,
1135.
(9) (a) Komander, D.; Clague, M. J.; Urbe, S. Nat. Rev. Mol. Cell Biol.
2009, 10, 550. (b) Haj-Yahya, M.; Eltarteer, N.; Ohayon, S.; Shema,
E.; Kotler, E.; Oren, M.; Brik, A. Angew. Chem., Int. Ed. 2012, 51,
11535. (c) Haj-Yahya, N.; Haj-Yahya, M.; Castaneda, C. A.; Spasser,
L.; Hemantha, H. P.; Jbara, M.; Penner, M.; Ciechanover, A.;
Fushman, D.; Brik, A. Angew. Chem., Int. Ed. 2013, 52, 11149.
(10) Akiyama, M.; Iesaki, K.; Katoh, A.; Shimizu, K. J. Chem. Soc.,
Perkin Trans. 1 1986, 851.
D
DOI: 10.1021/acs.orglett.5b01443
Org. Lett. XXXX, XXX, XXX−XXX