Overcoming the Hydrolytic Lability of a Reaction Intermediate in Production of Protein/Drug Conjugates: Conjugation of an Acyclic Nucleoside Phosphonate to a Model Carrier Protein

Article abstract of DOI:10.1021/acs.molpharmaceut.7b00410

Acquired immunodeficiency syndrome (AIDS) remains one of the most serious public health challenges and a significant cause of mortality for certain populations. Despite the large number of antiretrovirals developed in the past two decades, the inability of small molecule therapeutics to target HIV reservoirs directly creates a significant obstacle to their effective utilization. Indeed, achieving the desired therapeutic outcome in the absence of an effective means of targeted delivery must rely on dosage escalation, which frequently causes severe toxicity. This problem may be solved by conjugation of antiretroviral agents to endogenous proteins that are specifically recognized by HIV reservoirs (such as macrophages) for internalization and catabolism. However, conjugation of a large class of antiretroviral agents (acyclic nucleoside phosphonates, such as adefovir, tenofovir, and cidofovir) to a protein is challenging due to rapid degradation of the activated form of the drug (e.g., adefovir phosphonoimidazolide) in an aqueous environment. A novel synthetic strategy introduced in this work overcomes the instability of the activated form of adefovir by emulating the first step of its metabolic pathway (phosphorylation), making it highly reactive toward primary amine groups of proteins and, at the same time, less prone to hydrolysis by water. Efficient conjugation of the phosphorylated form of adefovir to a protein following activation with EDC (1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride) and imidazole was demonstrated using a model protein. Mass spectrometry (MS) was used to identify conditions that favor formation of conjugates with minimal side products, and online ion exchange chromatography/MS analysis of the products revealed the presence of multiple positional isomers within the 1:1 protein/drug conjugates. Both liquid chromatography/MS data and the analysis of ions produced upon top-down fragmentation of the 1:1 conjugates suggest that conjugation of phosphorylated adefovir to the protein occurs not only at primary amines but also at hydroxyl groups. The new conjugation protocol can be used to attach adefovir and other acyclic nucleoside phosphonates to proteins recognized by the cell surface receptors specific to macrophages (such as the haptoglobin/hemoglobin complex), enabling targeted drug delivery directly to HIV reservoirs.

Full text of DOI:10.1021/acs.molpharmaceut.7b00410

Subscriber access provided by University of Florida | Smathers Libraries
Article
Overcoming the hydrolytic lability of a reaction intermediate
in production of protein/drug conjugates: conjugation of an
acyclic nucleoside phosphonate to a model carrier protein
Shengsheng Xu, and Igor A. Kaltashov
Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.7b00410 • Publication Date (Web): 17 Jul 2017
Downloaded from http://pubs.acs.org on July 18, 2017
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.
Molecular Pharmaceutics 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 27
Molecular Pharmaceutics
1
2
3
4
5
Overcoming the hydrolytic lability of a reaction intermediate in production of protein/drug conjugates:
conjugation of an acyclic nucleoside phosphonate to a model carrier protein
6
7
Shengsheng Xu and Igor A. Kaltashovꢀ
8
9
Department of Chemistry, University of Massachusetts-Amherst, Amherst, MA, USA
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
ꢀ address correspondence to:
Igor A. Kaltashov
Department of Chemistry
University of Massachusetts-Amherst
240 Thatcher Way
Life Science Laboratories N369
Amherst, MA 01003
Tel: (413) 545-1460
Email: Kaltashov@chem.umass.edu
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 2 of 27
1
2
3
4
ABSTRACT
5
6
7
8
Acquired immunodeficiency syndrome (AIDS) remains one of the most serious public health challenges and
a significant cause of mortality for certain populations. Despite the large number of antiretrovirals
developed in the past two decades, the inability of small molecule therapeutics to target HIV reservoirs
directly creates a significant obstacle to their effective utilization. Indeed, achieving the desired therapeutic
outcome in the absence of an effective means of targeted delivery must rely on dosage escalation, which
frequently causes severe toxicity. This problem may be solved by conjugation of antiretroviral agents to
endogenous proteins that are specifically recognized by HIV reservoirs (such as macrophages) for
internalization and catabolism. However, conjugation of a large class of antiretroviral agents (acyclic
nucleoside phosphonates, such as adefovir, tenofovir and cidofovir) to a protein is challenging due to rapid
degradation of the activated form of the drug (e.g., adefovir phosphonoimidazolide) in an aqueous
environment. A novel synthetic strategy introduced in this work overcomes the instability of the
activated form of adefovir by emulating the first step of its metabolic pathway (phosphorylation), making
it highly reactive towards primary amine groups of proteins, and at the same time less prone to
hydrolysis by water. Efficient conjugation of the phosphorylated form of adefovir to a protein following
activation with EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride) and imidazole was
demonstrated using a model protein. Mass spectrometry (MS) was used to identify conditions that favor
formation of conjugates with minimal side products, and on-line ion exchange chromatography/MS
analysis of the products revealed the presence of multiple positional isomers within the 1:1 protein/drug
conjugates. Both LC/MS data and the analysis of ions produced upon top-down fragmentation of the 1:1
conjugates suggest that conjugation of phosphorylated adefovir to the protein occurs not only at primary
amines, but also at hydroxyl groups. The new conjugation protocol can be used to attach adefovir and
other acyclic nucleoside phosphonates to proteins recognized by the cell surface receptors specific to
macrophages (such as the haptoglobin/hemoglobin complex), enabling targeted drug delivery directly to
HIV reservoirs.
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
KEYWORDS
Human immunodeficiency virus (HIV), antiretroviral therapy, adefovir, protein-drug conjugate,
electrospray ionization mass spectrometry (ESI MS), ion exchange chromatography, LC/MS, top-down
MS/MS
ACS Paragon Plus Environment

Page 3 of 27
Molecular Pharmaceutics
1
2
3
4
INTRODUCTION
5
6
7
8
Acquired immunodeficiency syndrome (AIDS) remains one of the world’s most significant public health
challenges and a significant cause of death for certain populations.1 Despite the relentless search for a cure
of HIV infection, presently antiretroviral therapy (ART) remains the only feasible therapeutic modality that
allows many patients to achieve and maintain near complete suppression of the virus for extended periods
of time.2 However, the success of ART is undermined by the drug resistance,3 particularly in pediatric
populations,4 placing a premium on the continuous expansion of the repertoire of available ART agents.5
Unfortunately, a significant fraction of novel antiretroviral drugs fail to deliver on their promise, as the
inability of small molecule therapeutics to target the HIV reservoirs creates a significant obstacle to their
effective utilization. Indeed, achieving a desired therapeutic effect in the absence of an effective means of
targeted delivery frequently relies on dosage escalation, causing severe toxicity. For example, an acyclic
nucleotide analog adefovir6 that was initially developed as anti-HIV therapeutic failed to gain approval for
that indication due to significant nephrotoxicity, despite having demonstrated a clear antiretroviral benefit.7
Pharmacokinetic studies demonstrated that over 98% of unmodified adefovir molecules are excreted with
urine during the 24-hour window following an intravenous injection,8 clearly preventing it from reaching
anatomical and cellular viral reservoirs.9
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
Far from being unique to adefovir treatment, nephrotoxicity is a common problem for a number of
antiretroviral agents,10-13 imposing significant limitations on available therapeutic options. The problem of
dosage-dependent toxicity can be alleviated by targeting the ART agents directly to viral reservoirs,14, 15 and
limiting their levels elsewhere in the organism. One particularly attractive opportunity in this regard is
presented by haptoglobin, an acute phase plasma protein that irreversibly binds free hemoglobin molecules
in circulation and delivers them to macrophages for catabolism.16 Internalization of the
haptoglobin/hemoglobin complexes is mediated by CD163 receptors that are specific to macrophages and
other cells of monocyte lineage17 that harbor HIV virus in AIDS patients. The potential of CD163 to serve as a
target molecule for cell-directed therapies has been recognized several years ago.18 In fact, conjugation of
a nucleoside analog ribavirin to hemoglobin had been explored for targeted delivery to CD163-expressing
cells19 with the goal of overcoming the dose-dependent toxicity. While this particular nucleoside analog
cannot be used in ART, a strategy can be envisioned where ART agents are conjugated to
haptoglobin/hemoglobin complex to achieve targeted delivery to macrophages via CD163-mediated
internalization to inhibit reverse transcription of HIV in infected immune cells during early stages of acute
infection.20, 21 A particularly attractive class of agents includes nucleotide reverse transcriptase inhibitors
(NtRTIs) incorporating structural features such as phosphonate groups that prevent them from being
processed by nucleases.22 Targeted delivery of such NtRTIs utilizing vehicles capable of traversing
physiological barriers (e.g., transferrin or anti-transferrin receptor antibodies23) can also solve the problem of
poor accessibility of viral pools in the central nervous system, where cells of monocyte/macrophage lineage
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 4 of 27
1
2
3
4
5
6
(e.g., microglia, perivascular macrophages and meningeal macrophages) harboring HIV are protected by
the blood brain barrier.24, 25
7
8
9
Conjugation of one of the members of the NtRTI family, cidofovir, to a serine/tyrosine dipeptide has been
recently demonstrated,26-28 where the phosphonate group of cidofovir formed phosphonoester bond
with the hydroxyl group of serine in N, N’-dimethylformamide (DMF). However, this conjugate undergoes
rapid hydrolysis in aqueous environment at physiological pH with a half-life below two hours,26, 27 which
may not provide an adequate time window for successful delivery of an ART agent to its target site.
Furthermore, most carrier proteins are likely to undergo unfolding in organic solvent (such as DMF),
causing aggregation and precipitation during the conjugation process. Therefore, it is highly desirable
that the carrier protein/drug conjugation reactions take place in an aqueous environment under
conditions that do not compromise the higher order structure of the protein. In this respect, a more
attractive scenario would include two steps: chemical activation of the anti-viral agent in organic solvent
followed by the transfer of the chemically active intermediate to an aqueous protein solution to produce
a conjugate. This scheme was used successfully to produce a ribavirin/hemoglobin conjugate19 by
activating ribavirin with carbonyldiimidazole (CDI) in DMF to generate a reactive phosphorimidazolide
intermediate. However, our attempts to adopt this strategy to conjugate NtRTIs (such as adefovir) to
proteins failed due to the extreme instability of the phosphonoimidazolide intermediate in aqueous
environments compared to its phosphorimidazolide counterpart.
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
In this work, a modified conjugation strategy is presented which overcomes the problem of the
phosphonoimidazolide intermediate instability by chemically modifying adefovir prior to its activation.
This initial modification emulates the first step in adefovir’s metabolic pathway (phosphorylation),
allowing a phosphorimidazolide intermediate to be produced that is reactive towards primary amines on
lysine side chains and at the N-terminus of the protein (and, to a lesser extent, hydroxyl groups on the
protein side chains), and at the same time shows remarkable stability vis-à-vis hydrolytic attacks by water.
The new conjugation protocol can be used to attach adefovir and other structurally related anti-viral
agents to proteins recognized by the cell surface receptors specific to macrophages (such as a
haptoglobin/hemoglobin complex), enabling targeted drug delivery directly to the cells harboring HIV
viruses.
ACS Paragon Plus Environment

Page 5 of 27
Molecular Pharmaceutics
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
Materials. Adefovir (9-(2-phosphonomethoxyethyl) adenine, PMEA) was purchased from AK Scientific, Inc.
(Union City, CA). Lysozyme (Lz), 1, 1’-carbonyldiimidazole (CDI), phosphoric acid (crystalline, 99.999%
metal trace basis) and imidazole were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). EDC
(1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride) was purchased from Pierce
Biotechnology (Rockland, IL). Triethylamine (TEA), N, N’-dimethylformamide (DMF) and anhydrous
diethyl ether were purchased from Fisher Scientific (Fair Lawn, NJ). All other solvents and reagents were
of analytical grade or higher.
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
Phosphorylation of adefovir (PMEA) and preparation of lysozyme/adefovir conjugates (Lz/PMEA
conjugates). Preparation of the Lz/PMEA conjugates is shown in Schemes 2-3. PMEA was converted to
PMEA phosphonoimidazolide (PMEA-Im) using a modification of the previously published procedure.29
Briefly, PMEA (15 mg, 0.054 mmol) powder was placed in a 2 mL glass vial. CDI (0.082 mmol, 1.5 equiv)
was dissolved in 400 μL DMF/TEA mixture (3:1, v:v). This DMF/TEA solution was stirred at room
temperature for 1.5 hours, followed by addition of 36 μL of 3.1 M solution of anhydrous phosphoric acid
(0.054 mmol, 1 equiv) in DMF. The phosphorylation was performed overnight at room temperature.
Afterwards, the reaction solution was placed to cold diethyl ether to precipitate adefovir
o
monophosphate (PMEAp). The cloudy diethyl ether was kept at -20 C for 2 hours followed by
centrifugation, after which the supernatant was decanted. The waxy precipitate at the bottom was
re-dissolved in 100 μL distilled water and divided equally in two parts, which were mixed with 400 μL 0.2
M imidazole buffer (pH 6.8) and 400 μL 0.2 M sodium bicarbonate buffer (pH 8.7), respectively. EDC
(0.027 mmol, 1 equiv) was added to both reaction buffers and imidazole (0.027 mmol, 1 equiv) was
added to the 0.2 M sodium bicarbonate buffer. Both reaction solutions were stirred at room temperature
for half an hour. A 50 μL aliquot of the 1 mM lysozyme stock solution was added to each of the reaction
buffers (450 μL), giving a final protein concentration of 100 μM. The final drug-to-protein molar ratios in
both solutions were around 540:1. Both reaction solutions were stirred at room temperature for 24
hours.
Characterization of PMEA derived small molecules. 1 μL aliquots of DMF solution at different stages of
reaction (including PMEA’s complete activation by CDI and PMEA’s phosphorylation) were diluted with
100 μL deionized water followed by analysis performed with a QStar-XL (AB-Sciex, Toronto, Canada)
hybrid quadrupole/time-of-flight mass spectrometer to determine masses of the reaction/hydrolysis
products.
Characterization of Lz/PMEA conjugates. Mass measurements of Lz/PMEA conjugates were carried out
with a QStar-XL (AB-Sciex, Toronto, Canada) hybrid quadrupole/time-of-flight mass spectrometer
followed their purification by buffer exchange in 100 mM ammonium acetate; ions were produced by
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 6 of 27
1
2
3
4
5
6
7
8
9
electrospray ionization (ESI) using a turbospray source. The conjugates were completely denatured
during separation/purification on a 2.1 mm × 50 mm BioSuite (Waters, Milford, MA) C18 column
(particle size of 3 μm and pore size of 100 Å) prior to MS based characterization. Fragmentation of
conjugate ions was carried out by selecting a precursor ion (m/z selection window 1627.0 ± 0.4 u)
followed by its collisional activation (50 eV collisional energy). Assessment of the conjugate
heterogeneity (profiling of positional isomers) was carried out by introducing an online ion-exchange
chromatography (IXC) step prior to ESI MS measurements using a method developed earlier in our
laboratory.30, 31 Briefly, IXC (cation exchange) was performed on Agilent 1100 (Agilent, Santa Clara, CA)
liquid chromatograph equipped with a 2.1 mm × 100 mm PolyCAT A (PolyLC Inc., Columbia, MD) weak
cation-exchange column (particle size of 5 μm and pore size of 300 Å). Ammonium acetate at pH 7.0 was
used for MS-compatible gradient elution (100 mM in mobile phase A and 1 M mobile phase B). The flow
rate was maintained at 0.1 mL/min, and the linear gradient ran from 0% to 60% mobile phase B during a
60 min time interval. A Synapt G2Si (Waters, Milford, MA) hybrid quadrupole/time-of-flight mass
spectrometer was used for on-line ESI MS detection.
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 7 of 27
Molecular Pharmaceutics
1
2
3
4
RESULTS AND DISCUSSIONS
5
6
7
8
Preparation of lysozyme/adefovir conjugates. Our initial attempt to conjugate adefovir to a model
protein (lysozyme) using the procedure designed for conjugating a chemically similar drug ribavirin to
hemoglobin19 failed to generate any detectable conjugates despite the presence of seven primary
amines (targets of the conjugation reaction) within the protein. MS analysis of the activated form of
adefovir produced in DMF yielded a mass spectrum with a single peak at m/z 322 (Figure 1) consistent
with the notion of the expected intermediate (adefovir phosphonoimidazolide, see Scheme 1) being
produced at nearly 100% yield. However, the subsequent conjugation reaction must take place in the
aqueous environment, and addition of even small amount of water to the DMF solution of adefovir
phosphonoimidazolide (ca. 2.5 % by volume) resulted in a dramatic alteration of the appearance of the
mass spectrum following a relatively short (15 min) incubation time period (blue trace in Figure 1).
Although the adefovir phosphonoimidazolide ion is still present in the spectrum, the most abundant
ionic signal corresponds to a species at m/z 527; a lower-abundance peak corresponding to unmodified
adefovir (m/z 272) is also present. The mass of ion at m/z 527 is consistent with the dimeric form of
adefovir (linked via pyrophosphonate bond). This species (structure shown in Scheme 1) is formed not
only as a result of exposure to water, but also during diethyl ether precipitation (data not shown), an
essential step required to remove unconsumed reagents and organic solvents from the chemically
activated adefovir prior to its conjugation to a protein.
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 re-appearance of the signal of unmodified adefovir (m/z 272) in the mass spectrum of its activated
form (adefovir phosphonoimidazolide) exposed to water is likely due to hydrolysis (Scheme 1) facilitated
by formation of the transition zwitterion. Indeed, the methylene bridge (-CH2-) in the phosphonoamide
moiety exerts an electron donating inductive effect on both phosphorus and nitrogen atoms, leading to
their basicity increase. As a result, the proton of the hydroxyl group attached to the phosphorus atom is
expected to migrate to the nearby basic nitrogen atom of the imidazole ring, forming a transition
zwitterion and making imidazole a good leaving group.32 Formation of this species is expected to occur in
the phosphonoimidazolide intermediate, with the imidazole group being rapidly replaced by water (or
another agent capable of a nucleophilic attack.32 This behavior is in contrast to that of
phosphorimidazolide, which had been demonstrated to be much more stable to hydrolytic nucleophilic
attacks,33 and is likely due to the high basicity of the phosphorus–linked nitrogen atom (unlike
phosphonoimidazolide).32
The hydrolyzed form (i.e. adefovir) of adefovir phosphonoimidazolide (PMEA-Im) is strongly nucleophilic
and can also attack the phosphorus atom in the zwitterionic form of PMEA-Im, leading to formation of a
pyrophosphonate-linked PMEA dimer (Scheme 1), following essentially the same mechanism as the one
proposed for the hydrolytic attack by a water molecule. The molecular weight of this dimeric product is
consistent with the mass of the most abundant ionic species in the mass spectrum of water-exposed
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 8 of 27
1
2
3
4
PMEA-Im (m/z 527 in Figure 1), lending further credibility to the proposed mechanism.
5
6
7
8
The anomalous basicity of the nitrogen atom can be attenuated via a variety of ways, and perhaps the
most straightforward approach is to modify α-carbon of phosphorus in PMEA with an electron
withdrawing group, such as trifluoromethyl.32 However, the structural change caused by this particular
modification is likely to interfere with the drug’s ability to interact with the HIV’s reverse transcriptase,
making this approach unacceptable. An alternative strategy adopted in this work relies on extending the
terminal phosphonate group of adefovir by converting it to adefovir monophosphate (PMEAp, see
Scheme 2), which endows the molecule with a functional group that can be readily exploited vis-à-vis
conjugation to a protein without raised a specter of deactivation and/or forming non-productive
intermediates. Indeed, it had been reported that nucleoside diphosphorimidazolide is less prone to
hydrolysis than nucleoside monophosphorimidazolide, at least in the case of adenosine
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
5’-diphosphorimidazolide
and
7-methylguanosine
5’-diphosphorimidazolide
and
their
monophosphorimidazolide counterparts.33 In order to determine if phosphorylation of adefovir will
make the activated form of this drug less susceptible to nucleophilic attack, we first converted PMEA to
the monophosphate form (an analog of deoxyadenosine-5’-diphosphate (dADP)) to enhance the stability
of the reaction intermediate. The activated form of PMEAp (phosphorimidazolide) can be readily
produced in aqueous solution using EDC (Scheme 3). This strategy not only stabilizes the reaction
intermediate in aqueous environment without introducing significant alteration to PMEA structure, but
is also expected to facilitate the conversion of PMEA to its active metabolite (PMEApp) in vivo by
eliminating the need for one of the two obligatory phosphorylation steps. Conversion of PMEA to PMEAp
prior to its activation (to an imidazolide form) should also protect it against a hydrolytic attack not only
prior to, but also during the conjugation to a protein. Indeed, phosphorylation of PMEA to PMEAp (an
analog of dADP)33 reduces the basicity of nitrogen in phosphoramide moiety because the introduced
oxygen atom at the α-position of terminal phosphorus inductively withdraws electrons from the nitrogen
atom.34 This inhibits the formation of a transition zwitterion and stabilizes the intermediate. PMEA
phosphorylation was carried out by adding anhydrous phosphoric acid DMF solution in PMEA-Im. The ESI
mass spectrum of reaction mixture acquired in the negative ion mode (black trace in Figure 1) indicates
that PMEAp (m/z=352) is the major product, which coexists with a hydrolytic product (m/z 272) and a
diphosphorylated product (at m/z=432).
directly precipitated in cold diethyl ether.
To transfer PMEAp from DMF to water, this compound was
PMEAp extracted from DMF was re-dissolved in imidazole-HCl aqueous buffer (see Materials and
Methods for more detail), and activated by EDC in the presence of the target protein. The activation
proceeds via an O-acylisourea intermediate, which is known to be rapidly attacked by imidazole groups,35
a trait that was exploited in our work to conjugate PMEA to a protein (Scheme 3). The “on-demand”
formation of PMEAp-Im in the presence of the protein greatly reduces non-productive reaction channels
ACS Paragon Plus Environment

Page 9 of 27
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
(e.g., hydrolysis of this intermediate by water molecules), while allowing the nucleophilic attacks by the
primary amines (lysine side chains) to take place, leading to the conjugate formation. While the
unconsumed EDC is usually quenched prior to mixing with protein solutions with 2-mercaptoethanol, we
chose to avoid this step, as the presence of this reagent in the protein solution would almost certainly
induce reduction and/or rearrangement (scrambling) of disulfide bonds. The downside of keeping EDC in
solution is the occurrence of side reactions, although they can be suppressed by selecting appropriate
reaction conditions (vide infra).
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
Characterization of lysozyme/adefovir conjugates. ESI mass spectra of Lz/PMEA conjugates prepared at
pH 6.8 and pH 8.7 are shown in Figure 2. The ionic species present in each mass spectrum display narrow
charge state distributions, and the average extent of multiple charging is consistent with compact
(natively oxidized) conformations in solution.36 Both spectra contain signals of intact (unmodified) Lz and
a 1:1 conjugate (calculated mass shift 335.2 Da). In addition, the mass spectrum of the conjugate
produced under mildly basic (pH 8.7) conditions contains ionic signals corresponding to conjugates with
2:1 stoichiometry, as well as a range of other products. The mass analysis of these modified Lz species
reveals two types of chemical modification (addition of 71.1 and 128.2 Da, consistent with reactions of
residual EDC with the protein molecule37), as well as their combinations with each other and with
Lz/PMEA conjugates. Specifically, modification of carboxylic groups of aspartic and/or glutamic acid
residues is likely to give rise to unstable O-acylisourea intermediates, which are then converted to stable
N-acylurea products via the intramolecular O-to-N acyl rearrangement in the presence of excess EDC38
(see Scheme S1 in Supplementary Material). Since the two nitrogen atoms in O-acylisourea
intermediate are linked with ethyl and 3-dimethylaminopropyl groups respectively, the rearrangement
may lead to one of the two possible N-acylurea products, depending on which nitrogen atom the acyl
group migrates to. The secondary amides in N-acylurea byproducts undergo base-catalyzed hydrolysis
and produce N-acyl-N-ethyl-carbamic acid and N-acyl-N-3-dimethylaminopropyl-carbamic acid
derivatives corresponding to the mass additions of 71.1 Da and 128.2 Da (see Scheme S1 in
Supplementary Material).
No dead-end products are observed in the mass spectrum of Lz/PMEA conjugates produced at pH 6.8,
indicating that the unstable O-acylisourea intermediates undergo quick hydrolysis prior to acyl
rearrangement at neutral pH. The relatively low yield of conjugation at pH 6.8 is likely due to the
significantly higher stability of the reaction intermediate PMEAp-Im, as PMEA’s phosphorylation prevents
formation of a zwitterionic transition state. As a result, PMEAp-Im undergoes slower hydrolysis and
requires a stronger nucleophile to attack it. Primary amines (e.g., ε-amines of lysine side chains) can act
as nucleophiles only if they are deprotonated. While increasing the solution pH might seem an easy
solution to this problem, it comes at a price of a dramatic rise of the abundance of dead-end and mixed
products. Therefore, even though both the yield of the conjugation reaction and the average drug
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 10 of 27
1
2
3
4
5
6
loading are noticeably lower when the reaction is carried out at neutral pH, these conditions are still
preferable to mildly basic pH, as the products are much less heterogeneous.
7
8
9
Although the mass spectrum of the Lz/PMEA conjugates produced at neutral pH reveals only a single
conjugate (in addition to the intact unmodified protein), the sample may contain several isobaric species
that differ from each other by the location of the conjugation site. In order to evaluate
micro-heterogeneity of this sample with respect to possible presence of several positional isomers, it
was analyzed by ion exchange chromatography (IXC) with on-line detection by ESI MS. This technique
was recently developed in our laboratory30 and successfully applied for characterization of samples as
complex as PEGylated glycoproteins subjected to stress conditions.31 The chromatogram of the products
of Lz/PMEA conjugation show a series of peaks, and the on-line MS detection provides evidence that
nearly all species eluting before 55 minutes have masses consistent with 1:1 conjugation stoichiometry,
which only a minor fraction of species representing 2:1 conjugates (Figure 3). Integration of the
intensities of the Lz/PMEA conjugates’ peaks in UV absorption chromatogram (not shown) suggests that
the yield of the conjugation reaction is 23%, which is reasonably close to the 27% yield obtained by
integrating ion peaks in the mass spectrum of the reaction mixture (Figure 2).
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
PMEAp conjugation to the protein at the lysine side chain not only eliminates the positive charges on the
ε-amine group of this residue, but also introduces negative charges (phosphonate and phosphate groups
of PMEAp). Although the total number of charges is the same for all isomers, the unique position of the
PMEAp group on the surface of the protein will have a profound effect on the charge density surface
distribution. The surface charge density pattern is an important determinant of the protein retention on
the ion exchange resin;39 therefore, it should not be surprising that IXC allow a number of isomeric
conjugates to be resolved (Figure 3). The most abundant 1:1 conjugate elutes at 49 min. and likely
represents the product of conjugation at the N-terminus of the protein (the pKa of the α-amine is slightly
lower than that of ε-amines, making it a more favorable site for the conjugation reaction compared to
the lysine side chains). Six lower-intensity peaks are also observed in the extracted ion chromatogram for
the 1:1 conjugate with ionic abundance ranging from 5% to 25% of the base peak. In addition, a few
minor peaks (abundance < 5% of the base peak) are present in the chromatogram of the 1:1 conjugate.
This observation is surprising, as the number of chromatographically resolved isobaric 1:1 conjugates
exceeds the total number of amines (six ε-amines at lysine side chains and one α-amine at the
N-terminus of the protein). It might be tempting to attribute this mismatch to the presence of
conjugates where the integrity of the protein higher order structure had been affected by the chemical
modification (leading to an altered surface charge pattern and, as a consequence, a change in the
retention time). Such large-scale conformational changes result in changes in the appearance of the
charge state distributions of protein ions in ESI mass spectra;36 however, the on-line ESI detection
provides no evidence of alteration of the charge state distributions of ionic species representing 1:1
ACS Paragon Plus Environment

Page 11 of 27
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
conjugates across the entire chromatogram. Therefore, it seems unlikely that conjugation triggers loss of
the protein native structure. A plausible explanation for the large number of the observed isobaric
species could be the occurrence of protein deamidation during the conjugation process. Deamidation is
a common non-enzymatic post-translational modification (PTM), which frequently (and inadvertently)
occurs during a variety of protein handling steps,40, 41 making it a serious nuisance in biopharmaceutical
analysis. Asparagines followed by either a glycine or serine residue are usually prone to deamidation,42
and lysozyme contains two such asparagine residues (N59S60 and N103G104). Deamidation results in a
mass increase of only 1 Da, which is nearly impossible to detect when the protein mass is large; however,
in our case a one-mass unit shift in isotopic distributions of the Lz/PMEA conjugates should be readily
detected because of the relatively low mass of the latter. No such shift was detected for either of the
species eluting prior to 55 min. in the chromatogram shown in Figure 3, suggesting that deamidation
does not take place under the conditions chosen to produce the conjugates (pH 6.8).
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
In order to understand what contributes to the unexpectedly large number of isomeric 1:1 Lz/PMEA
conjugates, collisionally activated dissociation (CAD) of the mass-selected Lz/PMEA ions was carried out
and the mass spectra of the resulting fragments (Figure 4) analyzed. CAD of the Lz/PMEA ion at +9
charge state (m/z 1627.6) gives rise to a relatively small number of fragment ions. Singly charged
fragment ions are observed in the low-m/z region of the mass spectrum (m/z 274 and 354) and the
complementary fragments at high m/z region (m/z 1597.3, 1786.8 and 1796.8). The most abundant of
the two low-mass fragments (m/z 274) corresponds to a (PMEA+H)+, produced by the cleavage of P-O
bond in the phosphate group of the conjugate. The complementary ion is represented in the CAD mass
spectrum with a high-intensity peak at m/z 1796.8 (charge state +8), and a fragment ion with the same
mass but different charge state (+9) is observed at m/z 1597.3 (corresponding to the neutral loss of
PMEA from the precursor ion). The presence of these fragment ions in the CAD mass spectrum of the
Lz/PMEA conjugates is not surprising, since the phosphoester bonds are known to be labile in the gas
phase and undergo dissociation prior to the cleavage of the polypeptide amide bonds.43 What is
surprising, however, is the presence of the second peak in the low-m/z region of the CAD mass spectrum
(a singly charged species at m/z 354). This mass corresponds to a (PMEAp+H)+ ion, which can only be
produced upon fragmentation of the Lz/PMEA conjugates if PMEAp is attached to the protein via
hydroxyl-bearing side chain, such as serine or threonine, as opposed to a “conventional” linkage via the
amino group of lysine side chains or the polypeptide N-terminus. Phosphoserine peptides are known to
undergo phosphate cleavages upon collisional activation to produce dehydroalanine via β-elimination43
and gas phase fragmentation of serine linked PMEAp probably follows the same mechanism, producing a
singly charged PMEAp.
The existence of the Lz/PMEA conjugates where the pro-drug is linked to the polypeptide chain via side
chains with hydroxyl-group not only explains the appearance of the CAD spectrum (Figure 4), but also
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 12 of 27
1
2
3
4
5
6
7
8
9
accounts for the anomalously high number of chromatographic peaks in the XIC for the 1:1 conjugate,
which exceeds the number of primary amine groups in the protein (Figure 3). In retrospect, conjugation
at Ser (and possibly Thr) residues using the amine-targeting chemistry presented in this work may not be
that surprising, because their side chains become nucleophilic upon deprotonation which is induced by
surrounding basic residues (e.g., lysine and arginine) and is probably solution pH independent, while
neutral pH limits the fraction of primary amine that can participate in the attack of activated form of
PMEAp. From the practical standpoint, the ability of hydroxyl groups to compete with primary amines as
conjugation sites may actually prove beneficial vis-à-vis designing affective means of targeted delivery to
macrophages, as no Ser and Thr residues of haptoglobin are known to participate in binding to either
hemoglobin or CD163 (the cell-surface receptor that enables internalization of the Hp/Hb complexes by
macrophages) via either hydrogen bonding or Van der Waals interaction, while five Lys residues (K262,
K291, K297, K345 and K379) are known to participate in such interactions with hemoglobin.44-46
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 significant degree of structural heterogeneity of the 1:1 protein/PMEA conjugates produced in this
work using a model protein provides important lessons vis-à-vis production of adefovir conjugates with
real carrier proteins (such as hemoglobin or haptoglobin). Indeed, a very large fraction of hemoglobin
(αβ dimer) and haptoglobin surface (14% and 13%, respectively) are involved in the binding interface
when the hemoglobin/haptoglobin complex is formed, and these regions incorporate six amino acid
residues bearing primary amines (K291, K297, K345 and K379 on Hp, and V1 and K99 on Hb-α), together
accounting for 11% of all surface primary amines in the entire Hp/Hb complex. In addition, several Ser
and Thr residues located in the vicinity of the binding interface may also participate in PMEAp
conjugation, although in the absence of human Hp crystal structure it is difficult to judge the extent of
their side chain solvent accessibility and the local electrostatic environment (the two most important
factors influencing their reactivity towards PMEAp-Im). Since neither hemoglobin nor haptoglobin alone
can effectively interact with the CD163 receptor (a critical step in internalization of these proteins by
macrophages),18 it appears that covalent attachment of PMEAp to any of these side chains will have a
strong negative impact on the effectiveness of targeted delivery by compromising the integrity/stability
of the hemoglobin/haptoglobin complex. In the absence of the effective chemical means to direct the
conjugation reaction away from these critical sites, it would seem prudent to produce the conjugates
using the hemoglobin/haptoglobin complex as a starting material, instead of either hemoglobin or
haptoglobin alone. Formation of a stable complex prior to the conjugation procedure will shield the
critical Lys, Ser and Thr side chains from the PMEAp-Im. In the absence of such “natural” protection
during the conjugation step, due care should be exercised to ensure the ability of the conjugates to
associate with the physiological partner thereby enabling the receptor recognition and conjugate
internalization by the targeted cells.
CONCLUSIONS
ACS Paragon Plus Environment

Page 13 of 27
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
In this work, we introduced a robust synthetic strategy to overcome the instability of adefovir
phosphonoimidazolide by emulating the first step of adefovir metabolic pathway and used this strategy
to demonstrate successful conjugation of this anti-viral drug with a model protein. The phosphorylation
of adefovir allows the conversion of α-position functional group of the terminal phosphorus from
methylene (-CH2-) to oxygen, efficiently inhibiting a formation of transition zwitterion, which frequently
occurs in phosphonoimidazolide based intermediates and causes their hydrolytic instability. Mass
spectrometry was used to monitor the outcomes of individual steps in the adefovir’s activation, allowing
us to understand why the classical conjugation chemistry failed in this drug. Use of mass spectrometry
was also critical in selecting the optimal conditions for the coupling of adefovir monophosphate to the
model protein, as the conditions favoring higher conjugation yield (and higher drug-to-protein loading
ratio) were shown to also lead to formation to a large amount of dead-end cross-links and mixed
products. A reasonable compromise between modest conjugation yields (27%) and product purity
(almost exclusively 1:1 conjugate) was found; however, the conjugate was shown to exhibit significant
structural micro-heterogeneity due to distribution of the conjugation sites among 12 protein site chains.
This exceeds the number of primary amines (the intended targets of the conjugation reaction),
suggesting involvement of other group(s) in conjugation. Fragmentation patterns of the conjugates
activated by collisions in the gas phase provide unequivocal evidence that primary alcohols (side chains
of Ser and Thr) also participate in conjugation reactions. The novel synthetic strategy developed in this
work can be applied to other anti-viral drugs (such as tenofovir, which exhibits a structural template
similar to adefovir, including the presence of a phosphonate group, as well as other members of acyclic
nucleoside phosphonate family). A work is currently underway in our laboratory to apply the new
synthetic strategy to conjugate adefovir to a hemoglobin/haptoglobin complex, enabling targeted
delivery of this drug directly to cellular reservoirs of HIV. Catabolic processing of the
hemoglobin-haptoglobin complexes occurs inside the lysosomes. These organelles contain a number of
enzymes allowing them to effectively degrade a variety of biomolecules, including proteins and nucleic
acids. In particular, cells of monocyte/macrophage lineage contain a specialized acid phosphatase
(tartrate-resistant acid phosphatase, sometimes also referred to as macrophage acid phosphatase or
type V acid phosphatase47), which not only hydrolyzes a wide range of phosphates, but also
dephosphorylates proteins.48 Perinuclear localization of the majority of lysosomes,49 as well as the
presence of nucleoside-specific porters50 are likely to facilitate the transfer of the nucleoside analogs
cleaved off their carriers (hemoglobin and/or haptoglobin) from lysosomes to their ultimate destination,
the cell nucleus.
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
ASSOCIATED CONTENT
Supporting Information
Proposed mechanism of formation of N-acyl-N-3-dimethylaminopropyl carbamic acid (NDCA) and
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 14 of 27
1
2
3
4
5
6
N-acyl-N-3-ethyl carbamic acid (NECA) derivatives in production of Lz/PMEA conjugates in pH 8.7
condition.
7
8
AUTHOR INFORMATION
9
Corresponding Author: Igor A. Kaltashov, Department of Chemistry, University of Massachusetts-
Amherst, 240 Thatcher Way, Life Science Laboratories N369, Amherst, MA 01003. Tel: (413) 545-1460;
email: kaltashov@chem.umass.edu
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
Notes: The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The authors are grateful to Dr. Khaja Muneeruddin (presently at University of Massachusetts Medical
School) for help with setting up online cation exchange chromatography mass spectrometry (carried out
in the Mass Spectrometry Center at UMass-Amherst), to Dr. Cedric E. Bobst (UMass-Amherst) for helpful
discussions.
ACS Paragon Plus Environment

Page 15 of 27
Molecular Pharmaceutics
1
2
3
4
Schemes
5
6
Scheme 1. Activation of PMEA with CDI in DMF (top), formation of PMEA-Im transition zwitterion
7
8
(middle) and formation of dimeric form of PMEA (via pyrophosphonate bond) (bottom).
9
Scheme 2. Phosphorylation of PMEA.
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
Scheme 3. Activation of PMEAp with EDC in aqueous solution (top), conjugation of PMEAp with primary
amine (middle) and hydroxyl (bottom) groups on protein surface.
Figure legends
Figure 1. Hydrolytic instability of PMEA-Im in the presence of H2O as exemplified by the ESI mass spectra
(negative ion mode) of this activated form of adefovir in DMF (brown trace) and DMF/H2O mixture,
97.5:2.5 v:v (blue). See Scheme 1 for structures of species whose peaks are labeled in the mass spectra.
The mass spectrum of the phosphorylated form of PMEA (PMEAp) in DMF is shown in black.
Figure 2. ESI mass spectra of the products of Lz conjugation with PMEAp carried out at pH 8.7 (red trace)
and 6.8 (blue trace). The inset shows a zoomed view of ions with charge state +9: the base peak at m/z
1590.4 corresponds to intact Lz; mass shift of 335.2 Da corresponds to conjugation of a single PMEAp to
the protein, while additions of 71.1, 128.2 and 199.3 Da represent Lz’s N-acyl-N-3-ethyl carbamic acid
derivative, N-acyl-N-3-dimethylaminopropyl carbamic acid derivative and their combination respectively.
Figure 3. Total ion chromatogram (black trace) of Lz/PMEA conjugates produced at pH 6.8 and extracted
ion chromatograms for the unmodified protein (blue), 1:1 conjugate (purple) and 2:1 conjugate (red)
plotted with different magnification factors (the m/z values indicate the ions selected for plotting each
XIC). The inset shows zoomed views of ion peaks at charge state +8 from the mass spectra averaged
across the elution windows as indicated on the panel (peaks labeled with stars correspond to sodium
adducts).
Figure 4. Mass spectrum of fragment ions produced by collisional activation of 1:1 Lz/PMEA conjugates
(m/z 1627.6). The diagrams in the inset show structure of observed fragment ions (fragments that can
originate from precursors where PMEAp is conjugated either to primary amines or alcohol hydroxyls are
labeled in red, while the fragments that can originate exclusively from precursors where PMEAp is
conjugated to alcohol hydroxyls are labeled in blue).
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 16 of 27
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
ACS Paragon Plus Environment

Page 17 of 27
Molecular Pharmaceutics
1
2
3
4
REFERENCES
5
6
7
1. Deeks, S. G.; Overbaugh, J.; Phillips, A.; Buchbinder, S. HIV infection. Nature reviews. Disease primers 2015, 1,
15035.
8
9
2. Deeks, S. G. Towards an HIV cure: a global scientific strategy. Nat Rev Immunol 2012, 12, (8), 607-614.
3. Clutter, D. S.; Jordan, M. R.; Bertagnolio, S.; Shafer, R. W. HIV-1 drug resistance and resistance testing.
Infection, genetics and evolution : journal of molecular epidemiology and evolutionary genetics in infectious
diseases 2016, 46, 292-307.
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. Boerma, R. S.; Sigaloff, K. C.; Akanmu, A. S.; Inzaule, S.; Boele van Hensbroek, M.; Rinke de Wit, T. F.; Calis, J. C.
Alarming increase in pretreatment HIV drug resistance in children living in sub-Saharan Africa: a systematic review
and meta-analysis. The Journal of antimicrobial chemotherapy 2016.
5. Zhan, P.; Pannecouque, C.; De Clercq, E.; Liu, X. Anti-HIV Drug Discovery and Development: Current
Innovations and Future Trends. Journal of Medicinal Chemistry 2016, 59, (7), 2849-2878.
6. Barditch-Crovo, P.; Toole, J.; Hendrix, C. W.; Cundy, K. C.; Ebeling, D.; Jaffe, H. S.; Lietman, P. S. Anti-human
immunodeficiency
virus
(HIV) activity, safety, and
pharmacokinetics
of adefovir dipivoxil
(9-[2-(bis-pivaloyloxymethyl)-phosphonylmethoxyethyl]adenine) in HIV-infected patients. J. Infect. Dis. 1997, 176,
(2), 406-413.
7. James, J. S. Adefovir: FDA committee votes against approval. Food and Drug Administration. AIDS Treat.
News 1999, 330, (No 330), 1-2.
8. Cundy, K. C.; Barditch-Crovo, P.; Walker, R. E.; Collier, A. C.; Ebeling, D.; Toole, J.; Jaffe, H. S. Clinical
pharmacokinetics of adefovir in human immunodeficiency virus type 1-infected patients. Antimicrob. Agents and
Chemother. 1995, 39, (11), 2401-5.
9. Gupta, U.; Jain, N. K. Non-polymeric nano-carriers in HIV/AIDS drug delivery and targeting. Adv. Drug Deliv.
Rev. 2010, 62, (4–5), 478-490.
10. Rho, M.; Perazella, M. A. Nephrotoxicity associated with antiretroviral therapy in HIV-infected patients.
Current drug safety 2007, 2, (2), 147-54.
11. Post, F. Adverse events: ART and the kidney: alterations in renal function and renal toxicity. Journal of the
International AIDS Society 2014, 17, (4 Suppl 3), 19513.
12. Wyatt, C. M. Antiretroviral therapy and the kidney. Topics in antiviral medicine 2014, 22, (3), 655-8.
13. Milburn, J.; Jones, R.; Levy, J. B. Renal effects of novel antiretroviral drugs. Nephrology, dialysis,
transplantation : official publication of the European Dialysis and Transplant Association - European Renal
Association 2016.
14. Edagwa, B. J.; Zhou, T.; McMillan, J. M.; Liu, X. M.; Gendelman, H. E. Development of HIV reservoir targeted
long acting nanoformulated antiretroviral therapies. Current medicinal chemistry 2014, 21, (36), 4186-98.
15. Ramana, L. N.; Anand, A. R.; Sethuraman, S.; Krishnan, U. M. Targeting strategies for delivery of anti-HIV
drugs. J Control Release 2014, 192, 271-83.
16. Nielsen, M. J.; Moestrup, S. K. Receptor targeting of hemoglobin mediated by the haptoglobins: roles
beyond heme scavenging. Blood 2009, 114, (4), 764-771.
17. Graversen, J. H.; Madsen, M.; Moestrup, S. K. CD163: a signal receptor scavenging haptoglobin-hemoglobin
complexes from plasma. The International Journal of Biochemistry & Cell Biology 2002, 34, (4), 309-314.
18. Van Gorp, H.; Delputte, P. L.; Nauwynck, H. J. Scavenger receptor CD163, a Jack-of-all-trades and potential
target for cell-directed therapy. Molecular Immunology 2010, 47, (7-8), 1650-1660.
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 18 of 27
1
2
3
4
5
6
7
8
9
19. Brookes, S.; Biessels, P.; Ng, N. F.; Woods, C.; Bell, D. N.; Adamson, G. Synthesis and characterization of a
hemoglobin-ribavirin conjugate for targeted drug delivery. Bioconjug Chem 2006, 17, (2), 530-7.
20. Clercq, E. D.; Holy, A. Acyclic nucleoside phosphonates: a key class of antiviral drugs. Nat. Rev. Drug Discov.
2005, 4, (11), 928-940.
21. Xu, S.; Kaltashov, I. A. Evaluation of Gallium as a Tracer of Exogenous Hemoglobin–Haptoglobin Complexes
for Targeted Drug Delivery Applications. J. Am. Soc. Mass Spectrom. 2016, 1-8.
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
22. Clercq, E. D. The design of drugs for HIV and HCV. Nat. Rev. Drug Discov. 2007, 6, (12), 1001-1018.
23. Pardridge, W. M. Re-engineering biopharmaceuticals for delivery to brain with molecular Trojan horses.
Bioconjugate Chem. 2008, 19, (7), 1327-1338.
24. Varatharajan, L.; Thomas, S. A. The transport of anti-HIV drugs across blood–CNS interfaces: Summary of
current knowledge and recommendations for further research. Antiviral Research 2009, 82, (2), A99-A109.
25. Kumar, A.; Abbas, W.; Herbein, G. HIV-1 Latency in Monocytes/Macrophages. Viruses 2014, 6, (4),
1837-1860.
26. Zakharova, V. M.; Serpi, M.; Krylov, I. S.; Peterson, L. W.; Breitenbach, J. M.; Borysko, K. Z.; Drach, J. C.; Collins,
M.;
Hilfinger,
J.
M.;
Kashemirov,
B.
A.;
McKenna,
C.
E.
Tyrosine-Based
1-(S)-[3-Hydroxy-2-(phosphonomethoxy)propyl]cytosine and -adenine ((S)-HPMPC and (S)-HPMPA) Prodrugs:
Synthesis, Stability, Antiviral Activity, and in Vivo Transport Studies. J. Med. Chem. 2011, 54, (16), 5680-5693.
27. Eriksson, U.; Peterson, L. W.; Kashemirov, B. A.; Hilfinger, J. M.; Drach, J. C.; Borysko, K. Z.; Breitenbach, J. M.;
Kim, J. S.; Mitchell, S.; Kijek, P.; McKenna, C. E. Serine Peptide Phosphoester Prodrugs of Cyclic Cidofovir:
Synthesis, Transport, and Antiviral Activity. Mol. Pharmaceutics 2008, 5, (4), 598-609.
28. Plosker, G. L.; Noble, S. Cidofovir: A Review of its Use in Cytomegalovirus Retinitis in Patients with AIDS.
Drugs 1999, 58, (2), 325-345.
29. Brookes, S.; Biessels, P.; Ng, N. F. L.; Woods, C.; Bell, D. N.; Adamson, G. Synthesis and Characterization of a
Hemoglobin−Ribavirin Conjugate for Targeted Drug Delivery. Bioconjugate Chem. 2006, 17, (2), 530-537.
30. Muneeruddin, K.; Nazzaro, M.; Kaltashov, I. A. Characterization of Intact Protein Conjugates and
Biopharmaceuticals Using Ion-Exchange Chromatography with Online Detection by Native Electrospray Ionization
Mass Spectrometry and Top-Down Tandem Mass Spectrometry. Anal. Chem. 2015, 87, (19), 10138-10145.
31. Muneeruddin, K.; Bobst, C. E.; Frenkel, R.; Houde, D.; Turyan, I.; Sosic, Z.; Kaltashov, I. A. Characterization of
a PEGylated protein therapeutic by ion exchange chromatography with on-line detection by native ESI MS and
MS/MS. Analyst 2017, 142, (2), 336-344.
32. de Medina, P.; Ingrassia, L. S.; Mulliez, M. E. Synthesis of the First Stable Phosphonamide Transition State
Analogue. J. Org. Chem. 2003, 68, (22), 8424-8430.
33. Sawai, H.; Wakai, H.; Nakamura-Ozaki, A. Synthesis and Reactions of Nucleoside 5‘-Diphosphate Imidazolide.
A Nonenzymatic Capping Agent for 5‘-Monophosphorylated Oligoribonucleotides in Aqueous Solution. J. Org.
Chem. 1999, 64, (16), 5836-5840.
34. Engel, R. Phosphonates as analogues of natural phosphates. Chem. Rev. 1977, 77, (3), 349-367.
35. Hermanson, G. T., Chapter 27 - Nucleic Acid and Oligonucleotide Modification and Conjugation. In
Bioconjugate Techniques (Second Edition), Hermanson, G. T., Ed. Academic Press: New York, 2008; pp 969-1002.
36. Kaltashov, I. A.; Abzalimov, R. R. Do ionic charges in ESI MS provide useful information on macromolecular
structure? J. Am. Soc. Mass Spectrom. 2008, 19, (9), 1239-1246.
37. Hermanson, G. T., Chapter 3 - Zero-Length Crosslinkers. In Bioconjugate Techniques (Second Edition),
ACS Paragon Plus Environment

Page 19 of 27
Molecular Pharmaceutics
1
2
3
4
Hermanson, G. T., Ed. Academic Press: New York, 2008; pp 213-233.
38. Nakajima, N.; Ikada, Y. Mechanism of Amide Formation by Carbodiimide for Bioconjugation in Aqueous
Media. Bioconjugate Chem. 1995, 6, (1), 123-130.
5
6
7
8
39. Abzalimov, R. R.; Frimpong, A.; Kaltashov, I. A. Structural characterization of protein-polymer conjugates. I.
Assessing heterogeneity of a small PEGylated protein and mapping conjugation sites using ion exchange
chromatography and top-down tandem mass spectrometry. Int. J. Mass Spectrom. 2012, 312, 135-143.
40. Li, X.; Cournoyer, J. J.; Lin, C.; O'Connor, P. B. Use of 18O labels to monitor deamidation during protein and
peptide sample processing. J. Am. Soc. Mass Spectrom. 2008, 19, (6), 855-64.
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
41. Wang, S.; Bobst, C. E.; Kaltashov, I. A. Pitfalls in protein quantitation using acid-catalyzed O18 labeling:
Hydrolysis-driven deamidation. Anal. Chem. 2011, 83, (18), 7227-7232.
42. Robinson, N. E.; Robinson, A. B., Molecular clocks: Deamidation of asparaginyl and glutaminyl residues in
peptides and proteins. Althouse Press: Cave Junction, OR, 2004.
43. DeGnore, J. P.; Qin, J. Fragmentation of phosphopeptides in an ion trap mass spectrometer. J. Am. Soc. Mass
Spectrom. 1998, 9, (11), 1175-88.
44. Polticelli, F.; Bocedi, A.; Minervini, G.; Ascenzi, P. Human haptoglobin structure and function – a molecular
modelling study. FEBS Journal 2008, 275, (22), 5648-5656.
45. Andersen, C. B.; Torvund-Jensen, M.; Nielsen, M. J.; de Oliveira, C. L.; Hersleth, H. P.; Andersen, N. H.;
Pedersen, J. S.; Andersen, G. R.; Moestrup, S. K. Structure of the haptoglobin-haemoglobin complex. Nature 2012,
489, (7416), 456-9.
46. Fatunmbi, O.; Abzalimov, R. R.; Savinov, S. N.; Gershenson, A.; Kaltashov, I. A. Interactions of Haptoglobin
with Monomeric Globin Species: Insights from Molecular Modeling and Native Electrospray Ionization Mass
Spectrometry. Biochemistry 2016, 55, (12), 1918-1928.
47. Moss, D. W.; Raymond, F. D.; Wile, D. B. Clinical and biological aspects of acid phosphatase. Crit. Rev. Clin.
Lab. Sci. 1995, 32, (4), 431-67.
48. Pisoni, R. L. Lysosomal nucleic acid and phosphate metabolism and related metabolic reactions. Subcell.
Biochem. 1996, 27, 295-330.
49. Appelqvist, H.; Waster, P.; Kagedal, K.; Ollinger, K. The lysosome: from waste bag to potential therapeutic
target. J. Mol. Cell Biol. 2013, 5, (4), 214-26.
50. Lloyd, J. B. Lysosome membrane permeability: implications for drug delivery. Adv Drug Deliv Rev 2000, 41,
(2), 189-200.
ACS Paragon Plus Environment

Molecular Pharmaceutics
Page 20 of 27
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
352
100
75
50
25
0
527
322
432
272
200
250
300
350
400
450
500 m/z
ACS Paragon Plus Environment
Xu and Kaltashov, Figure 1

Page 21 of 27
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
+9
100
75
50
25
0
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
+ 335.2 Da + 334.8 Da
+8
+10
1600
1620
1640
1660
1680 m/z
1250
1500
1750
2000
2250
m/z
ACS Paragon Plus Environment
Xu and Kaltashov, Figure 2

Molecular Pharmaceutics
Page 22 of 27
1
2
3
4
5
6
7
8
9
4.00
3.00
2.00
1.00
0.00
1.00
0.75
0.50
0.25
0.00
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
+ 335 Da
+ 335 Da
58-59 min
48.5-49.5 min 34-35 min
1780 1800 1820 1840 1860 1880 m/z
m/z 1789
m/z 1831 (x 2)
m/z 1873 (x 20)
30
35
40
45
50
55
60
elution time, min.
ACS Paragon Plus Environment
Xu and Kaltashov, Figure 3

Page 23 of 27
Molecular Pharmaceutics
1
2
3
4
5
6
7
8
9
NH2
N
N
274+
1796.88+
100
1627.6 (+9)
N
N
354+
1786.88+
O
O
O
O P O P
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
OH OH
80
60
40
20
0
NH2
NH2
N
274+
1796.88+
N
N
OH
N
O
O
H
N
O
P O P
OH OH
NH2
N
1597.39+ 2730
OH
N
N
N
m/z
1780
1790
1800
O
O
274.1
354
H
N
O
P O P
OH OH
200
600 800 1000
m/z
400
1200 1400 1600 1800
ACS Paragon Plus Environment
Xu and Kaltashov, Figure 4

Molecular Pharmaceutics
Page 24 of 27
PMEA acꢀvaꢀon with CDI
1
2
NH2
NH2
N
PMEA-Im
(m/z 322)
N
N
N
3
4
5
6
7
8
9
N
N
H
N
N
CDI, TEA
22oC
N
CO
+
+
2
N
O
O
O
P
:
O
O
N
P
N
OH
OH
10
11
12
13
14
Formaꢀon of PMEA-IM transiꢀon zwiꢁerion and hydrolysis
15
16
17
18
NH2
N
NH2
NH2
PMEA
(m/z 272)
N
N
N
N
N
N
19
N
N
H
N
20
H2O
N
N
N
+
21
22
23
24
25
26
27
28
N
O
O
O
P
O
P
O
N
O
N
P
:
OH
O
HN
N
O
:
O
:
H
Formaꢀon of PMEA dimer
29
30
31
32
NH2
N
N
33
34
35
NH2
NH2
NH2
PMEA dimer
(m/z 527)
NH2
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
N
N
N
N
36
O
NH2
N
O
37
38
39
40
41
42
43
44
45
46
47
48
49
P
:
O
N
N
N
N
N
O
P
O
:
O
:
OH
N
O
O
N
O
O
P
P
O
P HN
O
HN
:
O
:
OH
:
O
:
O
:
O
:
O
P
HO
O
ACS Paragon Plus Environment
Xu and Kaltashov, Scheme 1

Molecular Pharmaceutics
PaMgeo2n5oopf 2h7osphorylaꢀon of PMEA
1
2
3
4
5
6
7
8
NH2
N
NH2
PMEAp
(m/z 352)
N
N
N
N
N
N
H
N
anhydrous phosphoric acid
N
+
22oC
N
O
P
O
P
O
O
O
N
O
O
:
HN
P
O
OH OH
:
:
9
10
11
12
13
14
15
16
17
Diphosphorylaꢀon of PMEA
NH2
N
NH2
N
PMEApp
(m/z 432)
NH2
N
N
N
N
N
N
18
N
N
H
N
anhydrous phosphoric acid
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
CDI, TEA
22oC
N
N
22oC
+
O
O
N
O
P
O
O
P
O
O
P
O
P
N
O
O
P
OH
P N
:O
:
O
:
P
O
O
O
O
O
:
OH OH OH
OH OH
ACS Paragon Plus Environment
Xu and Kaltashov, Scheme 2

Molecular Pharmaceutics
Page 26 of 27
PMEAp acꢀvaꢀon with EDC and imidazole
NH2
NH2
1
2
3
4
5
6
7
8
9
N
N
N
N
N
N
EDC, imidazole
N
N
O
P
O
P
O
P
O
P
O
N
N
O
O
O
OH
OH OH
OH OH
10
11
12
13
14
PMEAp conjugaꢀon with primary amine groups from protein surface
15
16
NH2
NH2
17
18
19
NH2
N
N
N
N
N
N
20
21
22
23
24
25
26
27
28
29
30
N
N
O
P
O
O
P
O
P
O
O
N
N
P N
O
O
H
OH OH
OH OH
PMEAp conjugaꢀon with hydroxyl groups from protein surface
31
32
NH2
33
34
NH2
N
N
N
N
N
35
N
:
O
36
37
38
39
40
41
42
43
44
45
46
47
48
49
N
N
O
P
O
P
O
P
O
P
O
O
N
O
O
N
O
OH OH
OH OH
ACS Paragon Plus Environment
Xu and Kaltashov, Scheme 3

Page 27 of 27
Molecular Pharmaceutics
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
ToC graphic
80x43mm (300 x 300 DPI)
ACS Paragon Plus Environment